Faulty TR1, TR2, a little smoke and some questions

  • No one logged in.

03-Mar-2020 02:09 AM

JFB

JFB

Posts: 11

Hi,

After I successfully upgraded my 33-303 set using Dada kits, a friend of mine brought me his dead 303 amp.

I noticed two "overheating" marks : one on the PSU board, the second on the right channel board near L100/R124/R125
I upgraded the amp using the Dada kit and powered the unit (probably not very smart). TR107 started smoking on the right channel so I pulled the wire.

I then noticed the circuit path was open between Pin 6 and R125 (near the mark). I'm pretty sure it was like that before the upgrade. I fixed it.
I tested R124 and R125 but I can't seem to get a precise reading.

The following components are faulty :

Tr1, Tr2 (0 ohm between E and C)
MR105 and MR106


- What would be the replacements to order for these components ? : TR1 & 2, TR107, MR105, MR106, L100, R124, R125
- I guess I should replace the same components for both boards to keep everything balanced ?
- From your experience, is there other elements
I should look intobased on this issue ?

Any help would be most appreciated
Thank you !
  Reply

03-Mar-2020 03:28 AM

JFB

JFB

Posts: 11

I just read on other posts that the MJ15003 is a good replacement for TR1 & 2 so that is checked out I believe.

  Reply

04-Mar-2020 04:49 AM

EJP

EJP

Posts: 1742

Tr1/2: MJ15003G. Use new washers and new silicon grease, or new silicon washers if you have them. Clean the heatsink area well first and test for shorts from transistor case to heatsink. The only one that should be s/c here is the middle one for the PSU.

Tr107: BC549. You must rebias after replacing this.

MR105/6: 1N4148, 1N4001, etc. If you want to raise the current limit and you have replaced the power transistors as above, you can use two or even three 1N4148s in series in each position. Observe polarity.

L100: you shouldn't need to replace this, but Element14 item #1644284 will do.

R124/5: you shouldn't need to replace these either, but again Element14 item #2448945 will do.

What has happened here is that one of the power transistors has gone s/c, which they can do any time: this has blown the PCB track; and then the other power transistor has blown itself up trying to balance the output point. You can repair the PCB track with wire: use a good thick piece, as it carries all the speaker current, and run it all the way from one component joint all the way to the other one at the other end, not just across the gap. Solder right along its length as far as each piece of remaning track goes, and bend it to the same curve so it stays clear of everything else.

I would only replace the components on the faulty board until you have it working, then mirror on the other board if you like. Only the MJ15003s will make any kind of difference to the sound.

  Reply

05-Mar-2020 01:50 AM

JFB

JFB

Posts: 11

Dear EJP,

Thank you for your help ! I am a bit of a newbie.

Washers : The "TO-3 Silicone-Fiber Insulator Pad" are out of stock on the webstore so I'll see if I can salvage the originals. As for the grease, I ordered the "Heat-Sink Compound 25g Servisol", hope this will do.

TR107 : When you says "rebias", do you mean readjust RV101 ?

L100 : I had ordered the 7,0┬ÁH 6A Epcos which someone mentioned on another post. R124-125 : The "0,3R 5W Precision Dale" is the only 0,3 Ohms I found in the webstore so I assumed it was "the one". I won't change them if not necessary.

As for the rest, I will follow your advices and update this post. Thanks for the explanations on what happened in the circuit.

Best,

JFB

  Reply

05-Mar-2020 02:02 AM

EJP

EJP

Posts: 1742

Don't, whatever you do, re-use the original TO-3 washers. You must use new ones.

Yes, 're-bias' means 'readjust RV101'.

  Reply

06-Mar-2020 01:51 PM

JFB

JFB

Posts: 11

Noted. I will get new ones

Thanks EJP.

  Reply

23-Mar-2020 12:49 AM

JFB

JFB

Posts: 11

A quick update

- 4x MJ15003G with new washers and silicone compound. I cheched them all prior and tested for shorts after. All good.
- BC549 for Tr107 on the right channel. I also replaced the wire (to fix the open path mentioned above) with a longer one, streching from R125 to pin #6 as per EJB's recommendation.

Powered the unit with both boards disconnected to adjust the PSU. 86V and RV200 doesn't modulate anything. Read all the posts mentionning this issue. Found Tr201 to be faulty and changed it with a BC556. Now 67V and I'm almost celebrating ...

Right channel :

1 to 9 : 67V
1 to 5 : 1.86 V (!?)

Removed TR100 to 106 and tested them with a basic component tester. All OK except Tr104 which I replaced with a BC556. This didn't solve the issue.

I took the following pin measurments (to chassis ground):
1,2 : 67.2 V
3: 66.0 V
4,5,6 : 65.5 V
7 : -24.7 mV
8 : -24.9 mV
9 : -12.7 mV

I am considering removing every componenent one by one to test them. Any hints on the ones I should check first will be most welcome.

Thank you kindly.

  Reply

23-Mar-2020 01:53 AM

EJP

EJP

Posts: 1742

1,2: OK.

3: this is the base of an NPN transistor, which should be up to 0.5V above the emitter, in this case pin 4, which it is.

4,5,6: obviously not OK but given these values we can say that the pin 3 voltage is OK.

7: like 3, this is an NPN base, so it should be up to 0.5V above the emitter on pin 8, which it isn't.

8,9: these are grounds; 8 goes to the emitter and 9 goes to chassis somewhere.

So clearly Tr2 is off, which explains why its collector is so high, and therefore all of 4,5,6.

So either Tr106 or one of its associated resistors, or Tr104 and ditto, are faulty. Or else the track repair wasn't correct, but the voltage measurements seem to rule that out.

  Reply

25-Mar-2020 01:13 AM

JFB

JFB

Posts: 11

Again, thank you EJP. Your answer is extremely helpfull and formative. 

I will poke around these components with my multimeter and see if I can pin point a fault.

A quick question : Should I be worried about feeding 67V (i.e. : plugging the unit) on the board in its current malfunctionning state ? I'm always afraid I will ruin or burn out 3 components while trying to fix one.

Best,

  Reply

25-Mar-2020 03:44 AM

EJP

EJP

Posts: 1742

Good question. In general you need current-limiting when testing: either a lightbulb in series with the mains, 60W for a 303, or else a current-limited DC lab supply like I have. If the lamp glows or the current-limit indicator shows, you have a fault.

However poking around while in this state is rarely of any use, as the rail voltage is going to be down due to the current limiting, so the voltages marked on the schematic don't apply, and they don't just scale linearly: for example, if I apply 30VDC to pin 1 of a 303 board which is working and correctly adjusted, the voltage at pin 5 will only be around 10V.

Second, you already know you have a fault. In this case I would be checking all the resistors I mentioned: if you get a strange reading, measure with one end unsoldered, so that you can rule out in-circuit measurement effects. If they are all OK, I would probably just replace all of Tr103/4/5/6, and then proceed with a current-limited supply to see what happens. You can test all the transistors in-circuit, not forgetting Tr2 which may have opened, using the diode-test feature of a DVM. Google will tell you how.

[If I replace one driver transistor I always replace the other one, for balance, as I won't be using the original type.]

  Reply

26-Mar-2020 03:37 AM

JFB

JFB

Posts: 11

Thanks EJP.
I will update the results.


  Reply

30-Mar-2020 02:03 AM

JFB

JFB

Posts: 11

Dear EJP,

I thought I'd play a bit with the simulation I made while waiting for parts, trying different malfunction scenarios to see how it affected voltage. A number of them gave me a pin read-out similar to what I had (above).

Tr104 short E/B was the most likely as you pointed out that Tr2 base was off. I had replaced Tr104 with a BC556 but still had the issue. I then swapped TR104 between the two board (left channel was OK) with similar results (!?).

Adding to the mystery, the left board seemed to be happy with the swapped Tr104.

Note : A C/E or C/B short on TR106 would drop the voltage on pin 4,5,6 according to the simulation.

Here's what I did :

1) Removed Tr105 and powered the unit - 54V on pin 5 (+ Tr105 tests OK)

2) Removed Tr103 (Tr105 still out) and powered the unit - 47V on pin 5 (+ Tr103 tests OK)

3) Removed Tr104 and TR102 (both test OK)

4) Desoldered 1 side and tested all following resistor which (if off or shorted) could induce the above pin read-out based on the simulation : R102, R104, R110, R112, R113, R114, R130. All resostors fine.

I changed Tr104 for a BC556 (again) and put everything back*. Pin 5 at 46V which is within the range of Rv101. Adjusted it and all is good. Still not sure exactly what was wrong ...

* On my unit, the emitter pin of Tr105 happens to be very close to the heat sink base (the aluminum flat with a hole in it under Tr105 and Tr106). I bent it a little to make sure it would not come in contact with the flat and kept the heatsink off to see. I'm thinking this could have been part of the issue. That said, an E/C short on Tr105 alone (without Tr104 faulty) would jack Tr2 to 140W (!!) according to the simulation.

The unit has been playing for 3 hours now ... sounds pretty good. Will update if it starts smoking again.

Again, thank you for your help. I would still be poking around clueless without your tips.

Best,

  Reply