Hi folks,

I'm attempting to revive a Heathkit AR-15 Receiver, and am seeking assistance beyond where my dusty college-level electronics knowledge can take me.

I don't know the history of the unit, other than that my father purchased it a number of years ago in non-working condition and it has laid dormant ever since. Its symptoms? Upon powering it up its transformer hums as though under heavy load, and the internal fuse blows in a heartbeat or so.

My thoughts so far have been to check out the PSU first, and upon inspection I found a few leaking electrolytic caps. All are easily obtainable locally, except one. 500uF 75V. I've highlighted it in the attached schematic. The closest I can track down to this is a 470uF cap with a 20% tolerance. 470 ±20% has a range of 376-564uF which strikes me as a bit low on the low end. Can anyone tell me whether or not this might be a suitable replacement, or would I be better off running a number of smaller caps in parallel to hit the target 500uF?


Thanks,

-d.

Welcome to AK,

Replace this cap by a 1000uF ~ 2000uF/75V (minimum, 100V preferable). It will be better. This is the first filtering stage in this power supply.

Also, check the diodes and power transformer primary.

Replace this cap by a 1000uF ~ 2000uF/75V (minimum, 100V preferable). It will be better. This is the first filtering stage in this power supply.

Will do.


Also, check the diodes and power transformer primary.

Added to my to-do list. The diodes in the bridge rectifier checked out (in-circuit), but that was about when I noticed the leaky caps and I wandered off in search of replacement parts at that point. :)

Overdue update...

The replacement caps were installed a couple of weekends ago, and I find it mildly amusing seeing modern, small-ish components installed on a board along side of relatively bulky 40 year old parts. :)

The transformer's primary and secondary windings checked out, however I found two bad diodes when I tested the bridge rectifier components out of circuit. Replaced 'em with a couple of spares, powered-up and the fuse still blew.

Pulled the bridge rectifier diodes again, and two more were shorted. Either something's killing old diodes on me, or I buggered up testing them... :scratch2:

Before powering up this time, I disconnected all of the outputs from the PSU board. Many power-ups and gradual re-connects later, I now have the unit powering up successfully with all of the components excecpt everything downstream of outputs 20 and 21. The isolated components at the moment are an 8k uF cap which is connected across ground and 20/21 through a pair of thermal circuit breakers, and the power amplifier boards.

The big cap appears to behave itself for resistance tests; my plan of attack at the moment is to see if she'll power up with only the cap connected, then move on to the power amp boards one channel at a time...

Terminal 20 on the PS board is rather an "input" for the high temperature control transistor. You should have the same voltage at pin 20 and 21 if the thermal breakers are closed (normal temperature). This turns off transistor Q503.

Remove the four output transistors, and test them. They are similar but take note of the actual position for each power transistor, just to see if other closely related components may be bad.

Terminal 20 on the PS board is rather an "input" for the high temperature control transistor.

Oops. :o I should have seen that. Transistor base != output!

You should have the same voltage at pin 20 and 21 if the thermal breakers are closed (normal temperature). This turns off transistor Q503.

So a lack of voltage at Q503's base would turn it on then (and in turn lighting the high-temp lamp)? I thought that transistors required a base current in order for them to close the emitter-collector path. :confused:

Remove the four output transistors, and test them. They are similar but take note of the actual position for each power transistor, just to see if other closely related components may be bad.

A multimeter test of a NPN transistor is similar to checking a pair of diodes whose anodes have been connected to each other, isn't it?

I'm looking for an open circuit from the emitter to the collector, and one-way current flow between the emitter/base and collector/base?

When there is a positive voltage (;17V) at Q503's base, the base emitter junction is reversed biased, so the transistor is off. When the base doesn't receive any voltage from the circuit breaker, current flows from emitter to base by the base resistance R513.

Basically, your transistor testing procedure is correct. The 40411 is a 90V, 30A, 150W with a hFE 35~100 at 4A

Output transistor Q206 for the left channel appears to have failed. There was bi-directional current flow between the base & collector, as well as uni-directional flow from the emitter to the collector. I had a replacement handy...


Pop goes the fuse. :scratch2:

Disconnect pins 2 and 5 from the power amp boards, the fuse will still blow. Heading upstream a bit further, I left C508 connected but broke the path through to to the output transistors... nope. Disconnect output 21 from the PSU board again, and the fuse still blows.

I'm out of time to pick at this for today, and I need to pick up another handful of fuses before I continue. Somehow it won't surprise me if I find another pair of shorted diodes in the bridge rectifier again. I'm going to pick up some replacements for those as well.

Does the PIV of diodes diminish with age? The replacements which I used earlier were from a parts amp of approximately the same age.

Disconnect output 21 from the PSU board again, and the fuse still blows.


I see two possible causes:

1) shorted rectifiers in the main bridge

2) shorted secondary. Disconnect one of the red wires from the transformer, and see if the fuse still blows. I'm crossing my fingers for you...

I don't think the PIV of diodes diminish with age, and a PIV of nearly 100V is not difficult to find.

The peak surge current in the bridge is high with a 8000uF cap, plus a shorted power transistor.

There was only one bad power transistor in the amp?

Today's efforts...

The transformer's 57v secondary is fine; the windings have a resistance of about 1ohm, and it produces a steady 60v when powered-up.

I found two more shorted-through diodes in the bridge. With those replaced and the 8000uF cap reconnected, the DC output is rock steady at about 1/2 a volt below spec.

Yes, I only found one bad output transistor last weekend, but I'm going to check them again today. I also intend to double-check the mounting of the pair labeled Q206 in the schematic I attached two posts earlier. The way they're mounted strikes me as being very easy for the collector (the transistors' casing) to become shorted to the heat sink. Onward! ......

Good news for the transformer.

Since there is no bias and no "offset" adjustments in this amp, I recommand that you test the amp without power transistors. Just to see if the voltage on the + side of the output capacitor is close to half the supply voltage. Power output are expensive, so if we can save a few in testing...Check Q205, this one may have been damaged by the shorted Q206.

The voltages on the schematic are not exact, it is impossible for the emitter of Q206 to be lower than the emitter if Q205.

Thanks for the continued coaching... :thmbsp:

All of the driver and output transistors check out, and the insulators between the the outputs' collectors and the heat sinks are in good shape.

Something I observed while testing the output transistors has my curiosity up though. As they're all the same type of transistor, I would have expected the base/collector and base/emitter resistance to be in the same ballpark. They're all over the place however...

Q206R
base/emitter 238k
base/collector 242k

Q207R
base/emitter 7.8M
base/collector 6.2M

Q206L
base/emitter 0.8M
base/collector 0.8M

Q207L
base/emitter 7.9M
base/collector 214k

Are these variances significant?


I'll continue with a test sans-output transistors tomorrow...

I imagine your "m" refers to megOhm. You should use capital M, since lower case m means milliOhm. This is confusing.

You can't test the transistors with your meter in this mode. From your measurements, I know you have a digital Ohmeter. When they measure resistance, they send a very very small current in the resistance, and they measure the voltage across the resistance. Then, they compute the resistance by division.

It is inapropriate for testing PN junctions, like you have in diodes and transistors. The voltage at the output of the meter (in Ohm mode) is too low for polarising the junction in conduction. The result is meaningless.

Your digital meter must have a "diode" check mode. In this mode, the voltage source is much higher. The reading will be in mV, like say 630mV if the probes are in one way, and "over" if the probes are the other way. This means the voltage drop in the junction is ~630mV when the junction is in conduction (forward biased), and the reverse breakdown voltage is beyond the limit of the voltage source inside the meter.

If your meter doesn't have this "diode" test mode, use an old analog Ohmeter. They work very well for testing PN junctions.

As an alternative, you can use a 1.5V battery and a limiting resistance (1kOhm for power transistors, and 10kOhm for small signal transistors should work) in series with the junction to test. Measure the voltage, with your meter in DC volt range, across the junction. You should measure ~600mV in forward biased junctions, and 1.5V in reversed biased junctions.

Managed to make some time today to get back to this... :)

I imagine your "m" refers to megOhm. You should use capital M, since lower case m means milliOhm. This is confusing.

Oops. The same rule applies in the PC world. I should have known that. :rolleyes:

You can't test the transistors with your meter in this mode. From your measurements, I know you have a digital Ohmeter. When they measure resistance, they send a very very small current in the resistance, and they measure the voltage across the resistance. Then, they compute the resistance by division.

It is inapropriate for testing PN junctions, like you have in diodes and transistors. The voltage at the output of the meter (in Ohm mode) is too low for polarising the junction in conduction. The result is meaningless.

Your digital meter must have a "diode" check mode. In this mode, the voltage source is much higher. The reading will be in mV, like say 630mV if the probes are in one way, and "over" if the probes are the other way. This means the voltage drop in the junction is ~630mV when the junction is in conduction (forward biased), and the reverse breakdown voltage is beyond the limit of the voltage source inside the meter.

Good to know... thanks. Using the proper mode on my DVM shows that all of the output transistors, including the one previously removed because I had thought it bad, are in fact ok.

C223 is a large can that might have dried out. Try taking this out of the circuit by disconnecting the wire going to it.

Then see whether you still blow fuses upon powering up.

Hope this helps ...

Regards,
Jerry

With the output transistors removed and pin 5 at each power amp board disconnected, there was still a short. This left only the trigger for the high-temp indicator circuit connected to the 80v supply.

It turns out that two wires had been swapped at the PSU board; pins 20 and 13. I'm not certain whether it was something which I had done while I had the board out for the capacitor replacement, or if it was a pre-existing condition, but with the inverted connections corrected the unit will power up nicely. Will replace the output transistors & reconnect the power amp boards tomorrow...

Well this thing's appetite for fuses appears to have finally been suppressed. :cool:

Next in line appears to be power amp issues. Other than a soft thump at power-up, there is no speaker output from any source. A hum injection test at the input of each power amp board also failed.

More exploration of the power amp boards to follow...

A hum test at the base of Q203 (pre-driver) passed, but the test at the base of Q201 (amplifier) failed. After a bit of exploring I discovered that the 27v supply appears to be dead.

Back to the power supply board... Voltages measured at all points downstream of the regulators checked out up to the "output" side of R501, which was at zero. When I attempted to remove the zener diode in this circuit from the board, without much prompting it fell apart. D'oh.

How are Zeners identified? Is it simply by their breakdown voltage, or are there other properties to be taken in to consideration? The broken diode has a graphic on it which possibly contains the letters "CT" or "ST", as well as the numbers 36 and 47.