Project: DIY More Tube-based Pre-Amp Kit (6J1 “Fever” Clone)

Ever since the vinyl “revival”, I’ve been bitten by a vintage audio bug of sorts. Of course, the old formats always come with their headaches and frustrations, but they also have a character and charm associated with them. One thing that’s always puzzled me is why people seem to have an affection for tube (or valve) based amplifiers. A lot is often said about the “warm” character of such amplifiers, which is merely a euphemism for some form of (potentially agreeable) harmonic distortion. As a person who has often been critical of audio compression and the (mis)use of graphic equalisers, it seems strange to me that people would seek to knowingly reproduce the audio in a less-than-ideal way.

Then it occurred to me that … before I tried vinyl records, I thought they were silly as well. After I tried them, I collected a small amount of them because the experience of putting on a vinyl record was rather amusing, but also because they didn’t sound half as bad as I had imagined. Maybe I should try a tube amplifier before I consign them to the pile of trash.

Of course, my budget (and power budget) really doesn’t stretch to buying a high-end Class A piece of Hi-Fi equipment – I don’t even use fancy speakers, often preferring my headphones. Instead, my thought was that perhaps a taste would be good enough. Many years back, I recall a “pre-processor” which had two tubes which you would run your audio through to give it the warm and fuzzy tube sound, despite using a regular transistor-based amplifier to actually drive the speakers. The unit was about AU$200 which was expensive in my eyes.

But a few years later, I spotted the DIY More kit which seemed to do the same thing, sold on eBay as a 6J1 tube-based pre-amp kit, a clone of the “fever” design which is very popular. The whole kit cost just AU$8.87 including the tubes, which I thought was a bargain, so I promptly ordered it and let it sit in my kit drawer.

Years passed, while it continued to sit in the drawer, without ever being assembled. Perhaps it was the fear that it was too complicated and I would screw it up … or perhaps the fear I would be wrong about my perception of tube-based amplifiers … but the day of assembly has now arrived.

The Kit

The familiar zip lock plastic bag makes a return, with all the components haphazardly thrown into it. I’m rather surprised at the lack of care given the fact they are shipping 6J1 tubes in the bag which are made of glass and have a fragile “nipple”, along with the ceramic tube bases … I was just fortunate that nothing broke in transit despite the lousy packaging.

The package consists of a choice of barrel jacks for power input, RCA plugs for input and output, a potentiometer to adjust the volume, a stack of Hyncdz capacitors (<sarcasm> what a premium brand! </sarcasm>), four axial capacitors, some diodes and resistors, two LEDs, four transistors and the two tubes. No instructions are included.

Luckily the silkscreening on the board is very comprehensive, allowing for easy construction despite some of the labelling being in Chinese. While this kit is branded DIY More, this PCB design is almost universal amongst the “Fever” 6J1 clones.

The double-sided PCB has green solder-resist on both sides and white silkscreening on both sides as well. The exception is the ground track which is uncovered – possibly to allow easier grounding with the chassis in a kit with an enclosure. Test points for filament voltage and the split voltage-doubled rails are clearly marked.

One concern I had was the two tubes I received were not identically marked. Ideally, in a stereo system with two channels, having them as evenly matched as possible is one of the aims and starting with a dissimilar pair of tubes is a bit disconcerting. I decided to build it anyway – this is probably just a consequence of the low price.

Assembly

Before starting, I always like to do a quick count of how many solder joints are necessary to complete the project, as that’s a good idea of how long it might take.

Barrel Power Jack - 3
2x RCA Jacks Assemblies - 8
2x Tube Holders - 14
9x Electrolytic Capacitors - 18
4x Axial Capacitors - 8
4x Transistors - 12
Dual Channel Potentiometer - 8
5x Diodes - 10
2x LEDs - 4
21x Resistors - 42
Total: 127 joints

As a result, this kit is probably among the more involved of the ones I have attempted thus far taking slightly more than an hour and a half to complete. As they are all through-hole components, it was easy to just populate and solder in almost any order. While the board has a HASL finish without any soldering-specific thermal management of the pads, the vias did make soldering a breeze and rather enjoyable.

This time around, I was running out of my preferred 0.8mm 60/40 solder with crystal clear flux and instead opted to complete the kit with a beefier 1.2mm 60/40 rosin core solder which leaves some flux residue. I didn’t bother cleaning it as the solder claims that it doesn’t need to be cleaned – however it does make the board slightly less neat in appearance.

It didn’t take too much effort and the board had all components populated. The need to mount the axial capacitors vertically was not implied by the silkscreen, but seemed to be the most sensible way to complete the kit. I suspect other types of capacitor may have been supplied with the other kits.

Then it was a case of installing the tubes which takes considerable effort and care to ensure you don’t break them and you get them fully seated. Unfortunately, I started mounting one of my tubes when I noticed it had a bent pin preventing it from properly seating. I had to extract the tube carefully, manually straighten and re-seat. It was a tense moment.

The other issue was that I had mounted the tube sockets not perfectly flush to the board, and as a result, the two tubes point slightly outward rather than being nice and straight. I guess that just goes to show that this one was “built at home”.

Testing

One of the biggest issues was to find a 12V AC/1A power supply, as many products now use switchmode regulated DC supplies due to efficiency regulations. I did manage to find one from some long decommissioned old telephony equipment, which I lopped the plug off and fitted a suitable plug.

It seems either the line voltage was high (it usually is) and the transformer output was high (quite probably) as the +/-28V rails were measuring 34V – just one volt shy of the capacitor’s maximum working voltage rating. I measured the current consumed with my clamp meter and it seems that the amp really only drew 390mA while driving a high-impedance input, so being below the 1A rating of the transformer probably led to the voltage floating upward somewhat.

The appearance of the kit is rather unusual – the blue glow from the LEDs under the tube bases is probably something I could do without as it obscures the orange-yellow glow from the heater filaments. Speaking of the heaters, the tubes come to life very quickly on applying power, taking just about 10 seconds to warm up.

I decided to hook it into my Rohde & Schwarz RTM3004 to see what it was like – the first test was just to see when the amplifier would be overdriven.

With an input (C2) of 5.5V peak-to-peak, on maximum gain, it’s clear that the output (C1) is clipping on the positive excursion about 12V above the ground level, whereas the negative excursion is easily reaching about -18V. I suspected I had an uneven voltage rail or built something wrong … but I couldn’t see my mistake and the rails were balanced when measured by a multimeter. I chalked this down to possibly the biasing point of the amplifier – I mean, you would not reasonably expect a line level input to be so “hot”.

Dialling back to a more appropriate line-level input of about 894mV peak-to-peak, the output achieves ~7.23V peak-to-peak. It’s rather smooth looking, with the tube amplifier being obviously an inverting amplifier. The left channel checks out …

… so I repeated it for the right channel. Despite the mis-matched tubes, the gain is roughly similar, with the maximum output of ~7.15V peak-to-peak for an input of 894mV peak-to-peak.

With shorted inputs, I was wondering what the noise output would look like. Driving a very high impedance, we see about 274mV of peak-to-peak noise, mostly 50Hz hum with some spikes. I wonder if some of that is being conducted into the amplifier in some way. This looks quite bad – but in reality when driving reasonable impedances (rather than an oscilloscope with 10:1 probe), it seems the hum falls quite dramatically.

As I have the Bode Plot option on the RTM3004, I decided to give it a shot and graph the frequency response throughout the complete range available. Rather interestingly, the unit has about 18dB of gain at 1kHz, dropping to 0dB at 982kHz which is very much into the AM-band … it seems like it’s probably quite a decent audio amplifier.

At maximum gain, the frequency response is quite remarkably flat, well within 0.5dB between 10Hz through to 24kHz. Left channel showed a gain of 18.39dB at 1kHz, dropping by 0.41dB by 24kHz. Right channel was 18.27dB and 0.42dB respectively, being fairly closely matched. Even the phase deviation of -12.42 and -12.79 degrees respectively is commendably similar, and still nowhere near danger of oscillation.

Configured for unity gain, it doesn’t seem that much is gained with very similar results – down by about 0.4dB at 24kHz with a phase deviation of -13 degrees. As a result, it doesn’t seem we are slew-rate limited in any way.

Putting the input into the left channel and plotting the bode from the right channel with the input shorted gives us a slight idea of what the cross-talk is like. The output is 18dB amplified, with the maximum cross-talk in the bass frequencies and upper treble. The output being -9dB to -18dB implies a channel-to-channel separation of about 27dB to 36dB which is sort of like FM-radio. At the best, it was around 50dB of separation. These figures are probably not the best, but may contribute to the character of the amplifier.

To explore how it actually sounded, I connected its input to my Asus Xonar Essence STX and the output into a Zoom H2n Handy Recorder. With this combination, I played a variety of music through the amplifier with it configured to amplify as much as possible while remaining below clipping on the H2n, reducing the output from the Xonar as necessary to keep the system happy. The recordings made at 96kHz/24-bit were then downloaded to the computer and listened to through the Xonar through my AudioTechnica ATH-M50x headphones.

From my experiment, I can definitely hear some sort of sound-stage altering effect which may be a product of the crosstalk which may be related to the design of the voltage-doubling power supply stage on the amplifier. This has the effect of having a slight cross-feed effect which didn’t seem objectionable. The main difference I heard was a slightly warmer vocal with increased mid-range oomph, but it was slight enough to be easily missed as using a different sound card/output op-amp. The amplifier didn’t seem to have any great negative traits such as hiss or noise aside from a hum that didn’t seem too severe and is to be expected from such a crude AC-power input stage and low-quality capacitors.

I actually began to like it somewhat – I didn’t feel like it was something I needed to have permanently attached to my system, but I didn’t find it to be anywhere near as bad as I had imagined it to be. It’s actually a rather competent little amplifier.

Conclusion

While I don’t think I could judge the idea of tube amplification from the results of an AU$8.87 kit, I was pleasantly surprised at how competent the kit was despite its low price, mismatched tubes and somewhat “hum-prone” design. The unit had 18dB of gain at the top end, channel-to-channel gain figures that were very close and an excellent frequency response. It may have been down on the channel-to-channel isolation and hum, but considering that it’s taking 12V AC and pushing that up by voltage-doubling to +/- 28V (nominal) and driving the filaments from half-wave rectified AC smoothed through a capacitor, I think that is to be expected. It’s a compromise which makes powering the kit simpler, not helped by the choice of cheap capacitors which may not have stellar ESR characteristics.

While the lowest-cost bare-bones kit may have been cheap, the packaging and shipping were a bit poor and it was a miracle that the tubes survived intact. That being said, without an enclosure included, it’s a little impractical to use as the tubes could easily be victim to a stray impact from an object, or the whole board might slide around on a table and expose potentially dangerous voltages (measured up to 68V non-ripple-free DC). Without an included 12V AC supply, it could be a little challenging to find an appropriate supply with the popularity of DC-output switch-mode supplies due to efficiency regulations.

On the downside, because it is so competent, the sound of the kit didn’t sound anything like I imagined. It wasn’t hissy, it wasn’t even “singing along” as I would have imagined. Instead, it seemed to sound like a crossfeed of certain frequency ranges and a slight increase in mid-range warmth with vocals sounding slightly fuller. Nothing that would jump out at me like silly EQ presets on a 90s stereo system.

In the end, I’ve got to say that I rather like the kit and how it turned out but I don’t think it’s something I would use everyday. I suppose I could always modify it to make it even better … maybe.

Posted in Audio, Electronics | Tagged , , , , , | Leave a comment

Tech Flashback: Anchor Automation Volksmodem 12

This weekend is turning out to be a bit of a telephony weekend, as I unpack more items from the stack of items donated by readers of the blog. This one is something I’m particularly fond of – the Anchor Automation Volksmodem 12 which is even older than the Netcomm SmartModem 2400SA I previously received.

In this post, I will look at the history of the company (Anchor Automation), the Volksmodem 12 itself, take a peek under the case and demonstrate it for your enjoyment.

Anchor Automation and the Rise of Modems

In the late 70s, dial-in bulletin-board systems (BBS) began to pop up, as the Hayes Smartmodem allowed for automated unattended placing of and answering data calls and computer visionaries started to write software to take advantage of these capabilities. This rapidly resulted in the development and growth of the modem business, as hosts felt the need to expand capacity through adding more lines and modems, and users began to be accustomed to the services, information and interactivity that BBS services could provide and thus purchased modems to connect their own microcomputers and PCs.

Unlike the internet, BBSes were more like a forum that was almost entirely text based where users interacted through a text terminal emulator to add postings, send messages, read news and later, even automate the exchange of files and play rudimentary games. These would usually operate in isolation, although some did allow some inter-BBS passing of messages or dial-through if you had the necessary permissions.

Born a little late to join the BBS era, it seems the early 80s was a boom period, as early 300 baud modems gave way to faster 1200 baud and then 2400 baud modems. Compatibility was always a concern, as modems were very expensive at the time and users were reluctant to upgrade or purchase something that wouldn’t give them a benefit for the servers they were dialling into.

The Hayes Smartmodem 1200 retailed for US$699 in 1982 (equivalent to US$1,830.99 in 2019) which was rather steep. Anchor Automation was a company that saw this as an opportunity to be a leader and introduce more cost-effective modems with the features and reliability demanded by users. The Volksmodem 12 was one of their answers – from this advertisement in InfoWorld in March 1985, they were retailing a comparable modem for just US$299 (equivalent to US$702.42 in 2019) with an industry-leading 5-year warranty. A look at another advertisement from InfoWorld in November 1984, Anchor Automation were pitching the Volksmodem 12 as “Porsche performance, Volksmodem price.” It seems the name, Volksmodem is inspired from Volkswagen – in essence, a modem for the people.

Indeed, the Volksmodem 12 was not their first product – the original Volksmodem was a 300bps modem, with the 12 being a “drop in upgrade” according to Computerworld October 1984. It would not be their last product either, with a higher end Signalman Express (300/1200bps) modem advertised with the tagline “Has the high cost of modems left your computer speechless? Smart Modem, Smart Price.”, the Signalman Mark XII (110/300/1200bps) modem advertised as “the Smart-Money Modem.”, Signalman LIGHTNINGi (2400bps), Anchor 1200-I/E and 2400-I/E, 24E5 and 2400PS. While their modems were cheaper than the competition, they were noted as having a few quirks, lower levels of FCC certification and rather average performance.

It would seem that in a growing market, Anchor Automation would be doing well, however, it seems that they were in decline since 1985 and were bought by PerfectData for $250,000 in a foreclosure sale in 1991. However, it seems that PerfectData didn’t have much of a strategy or direction, having a strong appetite for acquisitions with few plans. In 1995, they discontinued modem operations as their rivals produced lower-priced products with greater features, with the company being on the brink of collapse in 1998 when it pitched itself for acquisition. Digging further, it seems that PerfectData did make it to 2005 where they were acquired by Sona Mobile Inc. in a reverse merger, and in 2009, it seems Sona Mobile Holdings Corp. (SNMBQ) filed for liquidation in 2009.

The Volksmodem 12

The Volksmodem 12 has a rather plain, “brick” like monolithic appearance. Featuring a slightly yellowed beige exterior, it has a labelled front panel area with the branding clearly visible. The only user control is an on-off slide switch, with two LEDs for feedback indicating “high speed” (1200bps) and carrier detect. The modem comes with a grey fixed line cord terminating in an Australian 605 plug. This modem is likely to be from around 1986, so it’s probably something that belongs in a museum … actually – the Computer History Museum has this one and a Volks 6420.

The edges of the case are chamfered, but the sides of the case are enclosed with no ventilation holes.

From the rear, the line cord enters on the left cut-out with some space on the underneath. The right cut-out is patched over, seemingly modified from the original design which may have had two modular jacks. In the centre is a standard 2.1mm inner tip barrel jack for 12V DC input, positive centre. There is a 5-pin DIN data socket. A four-position DIP switch is mounted behind the hole on the right, with two switches visible for adjustment – I did not determine their function.

Luckily for me, a cable was included in the bundle, but the DB25 end has the wrong gender for fitting into an IBM-PC serial port. As a result, I had to use some creative use of a double-headed null-modem cable, plugging in the DB25 into the same short-end as a DE9 so I could get it connected to a modern USB-to-Serial adapter. Luckily, it worked – even with the tarnish on the DIN brass pins.

The pinout of the cable was measured as follows:

DIN  DB25  Description      _-- V--_  DIN Pin Order
1    2     TXD             /        \ Looking in from
2    7     GND            |          | pin side
3    3     RXD            | 1      3 |
4    8     DCD             \  4   5 /
5    6     DSR               -__2__-

The rear of the modem has four positions for feet which have since been lost – probably a good thing as most rubber feet disintegrate into “goo” sooner or later. This particular modem is Serial Number 566458 and seems to have been distributed by R.F. Computer Communication Pty Ltd. Unfortunately, this company too seems to have disappeared, according to the ASIC Commonwealth of Australia Gazette issued 10th May 1999, the company with ACN 006 279 189 was to be deregistered within two months of publication. I suspect they may have been responsible for any modifications to bring them into telecom authorisation for use in Australia.

Inside the Volksmodem 12

It seems the Volksmodem 12 may have been a little prescient – being a low-cost modem it uses a screwless design. That was not expected. The two halves of the case secure together by six friction-fit studs which have some silicone adhesive on them. As the adhesive had worn, opening the modem was as simple as prying the two halves apart under finger pressure.

A look inside shows a fairly neat double-sided PCB with green solder resist and white silkscreen. Marked Anchor Automation Inc. Assy 00672 REV, the chips on the board are dated from 1984 through to 1986. There is even a friendly LM317T and LM7805 linear voltage regulator we are quite familiar with. The brains of the modem is a Hitachi HD6301V1P 8-bit HD6801-compatible microcontroller with 4kB ROM, 128 byte RAM. The heart of the modem appears to be a Motorola MC14412 Universal Low-Speed Modem, Exar XR2120CP PSK Modem Filter and an Exar XR2209CP Voltage Controlled Oscillator. These are supported by a number of regular discrete logic and op-amps including MC1405x series Analog Multiplexer/Demultiplexers and an NE571 Compander.

The neatness, however, is broken by an add-on PCB that contains the phone line interface, patched into the main-board with a number of thin insulated wires.

The timing reference for the serial port seems to be a rather large 2.4576MHz crystal. There is a smaller 4MHz crystal nearby as well.

We can see the only things truly proud of the PCB are the two LEDs and the arm of the power switch. There also seems to be a trimpot which may adjust the carrier frequency of the modem.

A closer look at the line interface shows it is a very “hefty” old design using a traditional line hybrid transformer, relay for line switching and soldered cartridge fuse for protection. A small bit of foam which has since rotted away helps keep the heavy transformer from rattling too much inside the case.

We can see that when the line interface PCB is lifted (as it is not firmly secured to the board), it is insulated from the main board with a piece of card. Underneath, there appears to be a PCB design for a line interface involving the use of modular jacks, just that none of it is populated and the wires are “patched in” to pads almost at random. I suspect that the design may have been suitable for another country (e.g. USA) but does not meet Australian telecommunication standards.

It seems a rather old BFY50 NPN transistor in a metal can is also visible on a vertical mini PCB.

Being a two-sided PCB, there are traces on the rear as well, with the PCB dated Week 42 of 1986. The underside claims this board to be Revision C.

Getting it Running

Getting the modem running wasn’t too difficult. The first step was to plug the serial cable into my dual-headed null-modem cable as a gender changer and DB-25 to DE-9 converter. The DE-9 portion went into an Atten (Prolific PL2303-based) USB to Serial adapter, to be plugged into my computer. The telephone 605 plug was plugged into an adapter to modular jack, which was connected to my Cisco SPA112 ATA connected to my home Raspbx. Power was supplied from my Manson HCS-3102 at 12V/500mA current limited.

To have another modem to talk to, I chose my Rockwell/Conexant-based Netcomm Roadster II 56 Ultra SVD serial external modem as it was handy, attached to another USB to Serial adapter and attached to the second part of the SPA112 ATA. Traffic to and from the ATA was monitored by using port mirroring on my Netgear GS724T, captured by Wireshark and reassembled to produce the audio recordings.

On the whole, it worked almost as expected as soon as it was powered up, provided you used either 1200bps or 300bps line rate (no baud rate conversion is performed). The modem operates synchronously with the serial port rate and some subtle quirks with the AT command set and responses are noticed (e.g. there is no ati that reports the name of the modem). Auto-answer after one ring seems to be the default as well. Negotiations with modern modems are problematic, as the Volksmodem 12 has a rather short time-out which causes the connection to be aborted before the other modem has fallen back to a matching modulation.

Instead of harping on about the modem itself, I decided to let the modem speak for itself. I decided to write a short blog about the modem and transmit it modem-to-modem and make a video and audio recording of the result, calling on some ASCII art to bring back some nostalgia.

The 1200bps Self-Introduction

Attempting a 1200bps link to run this experiment did not prove too difficult, except for the modem’s sensitivity to VoIP disruptions, causing the link to drop mid-transmission. This may also be because the modem may have lost its tuning slightly and thus more easily “loses” the carrier.

In this case, text transmission was done from the Volksmodem 12 to the Netcomm with the call placed from the Volksmodem 12 to the Netcomm. A slight delay after connection is experienced as the modem waits to hear if the other end is going to use an error-correction mode, but times out. Some errors are visible, as is expected from non-error-corrected operation.

A look at the audio alone shows the nature of 1200bps mode, which is a PSK-based transmission.

As full-duplex echo-cancellation had not been perfected at this stage, the modem operates with two distinct carrier frequencies, one for each direction which result in data carriers which do not overlap. Thus, echoes from impedance mismatches from the outgoing transmission does not corrupt the incoming transmission. The audio file can be downloaded here.

The 300bps Self-Introduction

Attempting a 300bps version was much more challenging, as it did not seem possible to get the Volksmodem 12 to call the Netcomm and connect at 300bps reliably. Instead, the call had to be attempted in the reverse direction – from the Netcomm to the Volksmodem 12. As the Volksmodem had a 300bps serial link this time, it answered the call after the first ring at 300bps and a connection was made. Text transmission was made from the Netcomm to the Volksmodem, but despite numerous retries, it was not possible to sustain the link long enough to achieve the full transmission.

Despite this, it is still a good result, as I didn’t expect something of this vintage to even be workable. It really illustrates the speed, or lack thereof, of 300bps transmission. Likewise, two separate carrier frequencies are used, with FSK being the mode of modulation.

This particular mode has very harsh tones which could potentially cross-talk onto other lines, thus modem output power is probably limited and these modes were discouraged or even disabled in some later modems. The audio file can be downloaded here.

Conclusion

The Volksmodem 12, like the Volkswagen it was probably named after, was a cost-effective smartmodem from Anchor Automation that tried to offer the same features, reliability and compatibility with the Hayes SmartModems of the time. It looks rather basic and monolithic from the outside, with an odd-ball DIN connector for the serial data and modifications for Australian telecommunications compliance. This example was rather interesting as it had been well kept and still appears to function correctly even in 2019 (around 33 years old). It really shows us just how much bandwidth has increased over the years.

That being said, the fortunes of Anchor Automation were not great, seemingly in decline despite having several products on the market in the 2400bps era and being sold to PerfectData, who themselves wound up the modem operations in 1995 under pressure from other competitors.

Perhaps it would have been better to dig-out the Netcomm SmartModem 2400SA for the Volksmodem 12 to talk to as it would have been more “period-appropriate”, but I don’t have it to hand … so instead, it’s AT+MS (Modulation Select) to the rescue.

Continue reading

Posted in Audio, Computing, Tech Flashback, Telecommunications | Tagged , , , , , | Leave a comment

Teardown: Bosch DS835 TriTech PIR/Microwave Motion Detector

One of the things I bought home from the recent Wyong Field Day was a Bosch DS835 TriTech motion detector. While the box was beat-up, the unit inside looked clean and the seller wanted just $5 for it, so I decided to give it a new home.

The Unit

The box itself is quite scuffed, but there’s no mistake – this is a Bosch Security Systems DS835 TriTech PIR/Microwave Motion Detector. From the date code of 886, it appears to be made in June 2008 in China.

Inside, there is a copy of the installation instructions and the detector itself. The front face has the PIR window, an LED window and a blank spot where a logo badge would normally sit. Part of the reason I purchased this detector aside from the TriTech name was because a very similar model is installed in our own house.

This is where the DS835 model name comes in – that area for the badge is actually occupied by a logo for Detection Systems on our older sensors. The Detection Systems company was purchased by Bosch in 2000, but the sensor still retains its “DS”-model designator as it is reputable.

A label at the top illustrates FCC compliance …

… the underside has a little gap for a look-down zone, and a clip latch release hole to open the detector.

The rear of the detector unit has not been breached, but has a number of areas for cables to enter/exit the unit. The unit also has bevelled edges for mounting in the corner of two walls. From this, I can only assume the detector was never installed.

Teardown

To open the unit, it’s a simple matter of using a flat bladed screwdriver to release the latch and flip the unit open. The front cover shows a light-pipe for the LED indication and the PIR lens with many different segments. The lens itself can be masked from the inside, but isn’t adjustable otherwise.

Looking at the other side, it’s clear why – the top left corner has a black screw with a slot – to adjust the look-down angle, the whole PCB is slid up or down in the casing. Judging by the foam that is still inside, normally used to plug up the cable entry hole, it’s probably a brand-new sensor. We can see a limited amount of jumper-based configurability – sensitivity normal or intermediate and LED on/off. Microwave range is adjusted by a potentiometer.

There is the pyroelectric sensor which is responsible for the PIR segment, a microswitch that senses tampering with the case and the antenna for the microwave (~10GHz) radar. I suppose that could be where “TriTech” comes from – or maybe from the intelligent filtering algorithm that reduces false detections. A bi-colour LED provides feedback about which mode detected motion and whether it was tripped, with a self-test supervisory feature that reports with a blinking LED in case of microwave malfunction. Output is via a relay that is normally closed. The angled terminal blocks include some spare connections as well – but the tamper circuit and relay outputs are separate, with 12V (nominal) input on the left-most side.

Removing the whole PCB shows the rear is covered by a black plastic cover. This makes sense as vermin could cause inadvertent short circuiting between points on the PCB. There is a loop at the top, which I suspect is to hold the cable as it loops over to the terminal block.

It is possible to remove the remainder of the screws to remove the black plastic cover, but I wouldn’t suggest doing that …

It can be seen that the microwave radar section actually has a lot of similarities to a Ku satellite LNB – this is a classic dielectric resonator oscillator as evidenced by the brown “slug” of ceramic next to an microwave transistor.

Correspondingly, on the rear cover, we can see the screw that impinges onto this area to tune the frequency, as in an LNB to tune the local oscillator. Another interesting finding was the unit being powered by a PIC16C711 8-bit microcontroller, which has just 1K EPROM and 68 bytes of RAM, but other than that, there’s not much else to see aside from the LM324 quad op-amp.

Conclusion

While I knew what to expect from the insides of a regular PIR sensor, opening a multi-technology microwave radar-based sensor was interesting. I didn’t expect the paths of satellite LNBs to cross with motion detectors – it seems they use fundamentally similar dielectric resonator oscillator technology to create their ~10GHz oscillations which depends on high speed transistors and very precisely shaped PCB traces. The screw in the rear cover would be used to tune the oscillation frequency so that it would be at a regulatory-legal frequency.

It was also surprising to note that the PIR with all its touted intelligent filtering of false triggering runs on a Microchip PIC16C711 8-bit microcontroller, with the whole sensor consuming just tens of milliamps.

Posted in Electronics | Tagged , , | Leave a comment