1973 Magnavox. One dead channel.

[JunkForLess]

Re: 1973 Magnavox. One dead channel.

Okay, so the issue is that the part numbers on the transistors don't come up with anything on google. Thats why I was using the parts list to describe them. Here are the part numbers AS INSTALLED. Also, I attached pictures, forgive the messy board, it will get cleaned up once I figure out what's up.


Q16: 1522 BC3D

Q17: 1512 BC3D

Q21: 1512 BC3D

Q22: 1522 BC3D

[JunkForLess]

Re: 1973 Magnavox. One dead channel.

Q21 must be PNP. If C is 0V, B is 5V and E is .31V this looks like saturation and is a problem. This does not make sense as the transistor should be cutoff, seems it either has an E-C short or you have an NPN transistor installed here.

Well well well, so in diode test on Q21


C > B .116 forward

E > B .116 forward

E > C DEAD SHORT


I only tested them using C to B E to B and I did the same test but with the meter leads reversed. I don't recall testing E to C...

Going to pull working PNP from the other side and run the extra test.

Q17, shows the following:

C > B .113 forward

E > B .115 forward

E > C starts out as 0.117V forward then slowly keeps creeping up. It was up to 0.200 after about 10 seconds, then it starts to climb quicker.

[eccerr0r]

Re: 1973 Magnavox. One dead channel.

Interesting. Usually Si BJTs should be off or on and don't do the floaty thing. C-E should be open.

Also the voltages you're getting are super low, should be near 0.7V, my spare 2n3055 is like 0.55V. Did someone sub super leaky germanium parts in for the heck of it...?

Curious what's the parts list identification for D4/D5? You may need to swap this and other parts if they tailored it for germanium transistors.

[JunkForLess]

Re: 1973 Magnavox. One dead channel.

Interesting. Usually Si BJTs should be off or on and don't do the floaty thing. C-E should be open.

Also the voltages you're getting are super low, should be near 0.7V, my spare 2n3055 is like 0.55V. Did someone sub super leaky germanium parts in for the heck of it...?

Curious what's the parts list identification for D4/D5? You may need to swap this and other parts if they tailored it for germanium transistors.

NPNs have a forward of .510 or so. Maybe all 4 just need to be replaced? Who knows what abuse this saw during the disco era..

[JunkForLess]

Re: 1973 Magnavox. One dead channel.

I attached that section of the parts list.


So we may have found the big issue with this amp? What do you think?

If so, time to order parts and open a can of worms...

1) What capacitors does everyone like for the PSU and Audio sections? So far I came up with using Nichicon UHE in the PSU and maybe Nichicon UKZ in the audio? This is based off what is in stock at Digikey.

2) I saw someplace that its hard to get the replacement output transistors to bias right? Im a noob at this. Is that true? Can I buy the NTE's off digikey and have this thing work right? Or should I look for old stock name brand off Ebay?

3) The replacement resistors for R107, R101. The factory ones are.47ohm, Non-Inductive, ceramic blocks. I had a few from Amazon I used for testing, but I wish to replace them with something better, but now I noticed it seems to be a odd part? Should I just roll with the off brand ones for now?

Thanks for all the help with this.

[eccerr0r]

Re: 1973 Magnavox. One dead channel.

What's the drop across D5 now? It should be a plain silicon 1N4002 with a 0.7V drop according to parts list, and if it is a silicon diode then don't replace it if you're using a silicon MJ2955 for Q21 and 2N3055 for Q22 (replace both!). If it's a germanium you may need to swap it for a 1N4002 else you'll get distortion.

The circuit has no biasing pots and yes you will get some unwanted leakage and there's not much you can do about it unless you want to redesign the biasing.

Up to you what you want to do with the resistors. I'm no audiophile and would be happy with anything that would work there...

[JunkForLess]

Re: 1973 Magnavox. One dead channel.

What's the drop across D5 now? It should be a plain silicon 1N4002 with a 0.7V drop according to parts list, and if it is a silicon diode then don't replace it if you're using a silicon MJ2955 for Q21 and 2N3055 for Q22 (replace both!). If it's a germanium you may need to swap it for a 1N4002 else you'll get distortion.

The circuit has no biasing pots and yes you will get some unwanted leakage and there's not much you can do about it unless you want to redesign the biasing.

Up to you what you want to do with the resistors. I'm no audiophile and would be happy with anything that would work there...

D5 tests at .542v

I agree, I am looking for a simple repair at this point. Per the parts list, are the outputs silicon or germanium? The other smaller transistors are in the modern looking packages.

[eccerr0r]

Re: 1973 Magnavox. One dead channel.

parts list appears to be silicon though there are some things I'm not familiar with which may be related to germanium so that they could be swapped in when cheap.

BTW chances are you should remove R111 / R112 for the respective channel if you're replacing with silicon output transistors (which you may well need to, finding NOS NPN Ge transistors is probably hard, personally pretty much all my Ge transistors are PNP, and I think I may have just one NPN Ge transistor and it's not a power transistor). You could replace the resistor with a pot to be the bias adjustment though I'm not sure what's the best value. Probably less than 500 ohms but it may be touchy to adjust. On the other hand, if you leave the resistor in, you'll likely get distortion on the output -- and that, even as a non audiophile, I likely can hear as it would introduce odd harmonics. But TBH I've never had an amp with Germanium power output transistors, all the ones I've run across were already Si.

[JunkForLess]

Re: 1973 Magnavox. One dead channel.

parts list appears to be silicon though there are some things I'm not familiar with which may be related to germanium so that they could be swapped in when cheap.

BTW chances are you should remove R111 / R112 for the respective channel if you're replacing with silicon output transistors (which you may well need to, finding NOS NPN Ge transistors is probably hard, personally pretty much all my Ge transistors are PNP, and I think I may have just one NPN Ge transistor and it's not a power transistor). You could replace the resistor with a pot to be the bias adjustment though I'm not sure what's the best value. Probably less than 500 ohms but it may be touchy to adjust. On the other hand, if you leave the resistor in, you'll likely get distortion on the output -- and that, even as a non audiophile, I likely can hear as it would introduce odd harmonics. But TBH I've never had an amp with Germanium power output transistors, all the ones I've run across were already Si.

Well, I have a large selection of resistors. I also have a scope. If I leave the left channel with Ge transistors, could I use that channel as a reference to help tune the silicon replacements on the right channel? or will it never sound right unless I switch both to silicon and fiddle around?

Thanks.

[eccerr0r]

Re: 1973 Magnavox. One dead channel.

Depends on your ears. Well, at least after you get the bias correct.

One channel may sound slightly different than the other is the only concern, you may not be able to discern the difference with random music, at least I'm pretty sure I won't notice the difference.

If you measured 0.5V on (both) the NPN(s) you may already have Si transistors there which is quite odd based on the markings. Perhaps it would be an upgrade having both being true complementary pairs that have more similar characteristics...again as a non-audiophile I doubt I'd be able to hear the difference.

Either way it'd sound better than a monaural amplifier on stereo input...

[JunkForLess]

Re: 1973 Magnavox. One dead channel.

Thanks for all the info. I was actually excited that the console was broken when I bought it. I wanted a project I could learn a bit from.

I think I understand now. Again, forgive me for being so new. In the schematic, the forward bias voltage at Q21 and Q17 is a bit low at 300mV as where Q16 Q22 are at 500mV. Both should really be up near .700mV? With enough fuss, do you believe this can be corrected at D5/R112 like you said or will there need to be something else changed? For example, anything near R99/R100 or maybe R101/R107?

Thanks.

[eccerr0r]

Re: 1973 Magnavox. One dead channel.

The 700mV number is just a rule of thumb based on current flow and something above 500mV is fine for silicon. I was mostly using this number to distinguish Si from Ge, as the circuit may/is tuned for specific transistors and willy-nilly swapping will cause anything from awful audio to magic smoke leakage.

[JunkForLess]

Re: 1973 Magnavox. One dead channel.

How about NTE219MCP or NTE180MCP?

[eccerr0r]

Re: 1973 Magnavox. One dead channel.

Probably not worth to go for higher power transistors, better to deal with the bias/offset.

[JunkForLess]

Re: 1973 Magnavox. One dead channel.

Got it. I'll stick with: NTE130/2N3055 and NTE219/MJ2955

Thanks.

[eccerr0r]

Re: 1973 Magnavox. One dead channel.

Come to think of it, anyone here can find a part number for a Germanium NPN power transistor (3A Ic or can dissipate 6W or higher)... and one that's in a TO-3 (versus things like TO-36 or one of the stud mounts)?

Almost seems as if it's TO-3 and NPN it's guaranteed to be silicon?

[JunkForLess]

Re: 1973 Magnavox. One dead channel.

My parts list has alternatives for the outputs. One option it lists is a pair of HEP-247 and HEP-248. Check out this old page I found on google.

[eccerr0r]

Re: 1973 Magnavox. One dead channel.

Well HEP-247 and HEP-248, like others on that page, are all silicon...

It is kind of weird, seems hole mobility is bad in Si (hence HEP248 and likely MJ2955 are more expensive than HEP247/2N3055) but the issue is not as bad as other issues in Ge.

---
interesting, 2n3055 is more expensive than mj2955 on mouser... might be demand inflation.

[JunkForLess]

Re: 1973 Magnavox. One dead channel.

So the only item in stock was NTE219MCP, "NTE219MCP is a matched complementary pair containing 1 each of NTE219 (PNP) and NTE130 (NPN)."

So I ordered two sets. But I swear both sets test backwards? The NTE130 tests like a PNP and the NTE219 tests like a NPN... or the pinouts are wrong? or what is my brain malfunction here..

NTE219 (Philips ECG219):
+ B>C - .568
+ B>E - .568
+ C>B - 0
+ E>B - 0

NTE130: (Philips ECG130):
+ B > C - 0
+ B > E - 0
+ C > B - .553
+ E > B - .560

[eccerr0r]

Re: 1973 Magnavox. One dead channel.

At least they are complementary.

Compare with a known good PNP/NPN transistor you have or make a simple amplifier circuit to test. At first glance, yes that looks wrong, NTE219 should be PNP whose collector should be grounded, NTE130 should be NPN whose collector should be connected to V+.

Was about to try to rule out multimeters, had lots of old analog VOMs that connected pos and negative wrong, but DMMs with diode check all should be correct. If you have another multimeter, connect + to + and - to - and measure the voltage, it should be positive. Though this step should not be necessary...

And hope "0" you meant "OL" or "1" ...