The Heathkit AR-15 receiver was produced for two or three years starting in 1967. It was available both as a kit and pre-assembled, and was well thought of at the time. The design is almost entirely discrete transistors, except for two simple ICs in the AM/FM tuner board.

I bought mine used about 40 years ago, from (I think) the guy who built it in 1968. A few years ago, I realized the FM demultiplexor was not working, so FM stereo broadcasts were not played in stereo. Not a huge deal, but it bugged me.

I moved this receiver into the garage a few months ago, where it worked fine (who needs stereo in the garage?), until it stopped working - it got intermittant, and then quit entirely. So, onto the repair bench it went.

The problem was power supply failure, so time to figure out why.

Working on the individual boards in the AR-15 is a pain, because none of them have connectors to the wiring harness, and most of them have about 20 wires connected to them. The power supply board has 25. I elected to cut them all, extend them as needed, and insert a DB15 connector pair in circuit so I could easily disconnect and reconnect the board. This turned out to be the best mod in the world, because I had that board in and out dozens of times.

The power supply failure turned out to be due to a failed driver transistor for the 50v series pass regulator. Heathkit transistors in the AR-15 are often proprietary, so you just have to find a substitute based on voltage, current, and maybe gain (hFE). Anyway, I replaced the transistor, Q502, with one from my "junk box", a 2N2405, and now the AR-15 lived again.

Power Supply Repairs and Modifications

Since I had the board out, I tested the caps and the carbon comp resistors. Two of the electrolytics had failed, and several of the carbon comp resistors had drifted high (like 30% or more), which they are wont to do. I replaced all of that.

Working with the AR-15's power supply reminded me how crude the design is. It was reasonable for 1967, but almost 60 years later, we can do so much better. The original uses a two transistor series pass regulator to drop the filtered DC from the (discrete) diode bridge down to 50v. The 50v is used by the phono preamp, and also dropped down to 27v, 15v, 12.7v, and 12v for the rest of the low power circuitry. Therefore, the 50v regulator has to supply the current for all the low power circuits. The amounts to about 300mA. Hence the pass transistor is bolted to the chassis for a heat sink.

The regulators for the 27v and the 15v are just a big resistor and a zener diode. The regulator for the 12.7v is a big dropping resistor feeding a transistor regulator, and the 12v is derived from the 12.7v using a 2 Watt dropping resistor. The three big dropping resistors get VERY hot. They dissipate several watts each.

The 15v regulator (zener diode) is easy to replace with a LM7815. The 15v load is about 70mA. Eliminating the zener saves the additional current that the zener needs. The LM7815 will not handle an input over 35v, so it needs something to drop the 50v down below that. The stock Heath resistor is nominally a 270 Ω 7 watt resistor, and I'm still using that -- and it gets really hot - about 100° C.

The 12.7v regulator (pass transistor with resistive divider off the 15v output) was replaced by an LM317 regulator, with an appropriate resistive divider to give 12.7v. This value of this replacement is debatable, since there's no efficiency improvement. There is maybe an accuracy improvement. I left in the resistor dropper to feed the 12v. Total current out of the 12.7v regulator is 138mA.

Initially, I had planned to use a switching (buck) regulator to drop the 50v to 27v, and then regulate the 15v and 12.7v down from 27v, but when I tried this, I found the noise from the switching regulator was audible. Sanity prevailed, and I gave up on the switching regulator plan, so I left the 27v zener and dropping resistor in place. The stock dropping resistor is 390 Ω 5 watt, and it also gets super hot - 70° C.

Current plan is to replace all three big ceramic power resistors with aluminum bodied power resistors, and then bolt those to the bottom cover to help dissipate the heat.

By the way, in a prior repair (decades earlier) I replaced the under-heat-sinked rectifier diodes for the power amp final stage with an aluminum-block diode bridge. I cleaned up this wiring while I was in there, to minimize wiring to and from the power supply pcb.

FM Multiplex Repairs

Based on my experience finding bad caps and carbon comp resistors in the power supply, I decided it wouldn't hurt to check all the parts on the multiplex board. So I desoldered all 20-odd wires, and unbolted the multiplex board. You can usually check carbon comp resistors in circuit, because they drift high with time. So if you measure one, and it's high, it's "bad". Other surrounding circuitry does not affect this decision, because other circuitry can only lower the measured value; so if the value is high, the resistor must have a high value. How high is too high is up to you; I elected to replace resistors that were more than 10% high. (That's why "bad" is in quotes above.)

I also tested all the transistors, and found one bad one. Several small electrolytics had also gone bad, and I replaced all that. Finally, I was able to bolt the board back in, and re-connect all the wires. Before bolting the board in, I also added threaded inserts to the chassis, in order to replace the fiddly lock washers and nuts, because they were in a really tight spot that would be super annoying to re-install.

Finally, I was able to do a re-calibration of the multiplex board, and it worked! Stereo FM was back! That's one great thing about Heathkits -- the manual is comprehensive, and they always published a manual for all their kits (obviously).

Phono Preamp Repairs and Modifications

I left the phono preamp for last, partly because the board is all but hidden. But turning the volume up with no phono signal resulted in a lot of hiss, suggesting some of the transistors in the phono preamp had gone noisy (this can happen with old transistors.)

To get the phono preamp out, you first remove the dial cord pulley, and carefully set it aside so as to not disturb the dial pointer calibration. Then, remove four self tapping screws, and lift up the tuner board (it's hinged) and use the provided prop (thanks, Heath engineers) to lock it up out of the way. Then you can unscrew the phono board mounts, lift it up, and unsolder the nine (only!) wires.

I just guessed that the first of the two transistors in each channel were the noisy ones, and I was right. Replacing the MPS6520s (an actual part number you can look up) with 2N3904, which the Motorola book calls a "low noise audio" transistor. This substitution seems to work fine, and noise is vastly better.

Assembly, as they say, is the reverse of disassembly. Attach wires, bolt in board, replace tuner board, replace dial cord pulley.

The Rest of the Boards

The tuner board can be pivoted up (as described above) and then both sides reached in order to test/replace caps and carbon comp resistors.

The amplifier driver boards can be pivoted up by unsoldering just the two wires at the end nearest the front of the unit, and then both sides can be reached.

I didn't even think about trying to work on the preamp/tone control board. It's buried behind the front panel, and I just threw up my hands at the prospect of working on it.

For all the other boards, I found bad caps and drifted carbon comp resistors, so this is worth doing to as many boards as you have patience for on any AR-15 you want to save.

Additional Failures and Fixes

After I fixed the power supply the first time, but was experimenting with new power supply circuitry (trying to lower the waste heat), the big regulator transistor Q501 turned into a low value resistor. This device has a proprietary part number, but a 2N3055 is a good replacement.

Weeks later, after I thought I had this thing "all fixed", the FM receiver got wierd. It would play some stations, but not others; stronger signals would result in no sound! And then, stereo stopped working!

Fortunately, Heath's manual includes voltage readings for all the transistors in the unit, so I measured the voltages on the tuner board, and found a bad transistor (Q311, an MPS6517). I had an MPS6518 in stock, so once replaced, FM is back to normal now. (The MPS6518 is an MPS6517 selected for higher gain. Higher gain is not a problem in this application.)

Unrelated Product Recommendation

I bought a Milesseey TR10 IR camera based on a review on the YouTube "Project Farm" channel, and it's the best thing since sliced bread. 240x240 resolution, so sufficient for debugging electronics. Not terribly expensive; I paid < $150 (US). Recommended.

FAQ

Q. Will you fix mine?

A. No, I'm retired and so not interested in a job.

Disclaimer/Warning

This is just my documentation for my repair. I don't claim that doing this is safe or recommended.

Soldering irons are dangerous, be careful. Oh, and don't eat the solder.

William Dudley
June 27, 2026