Maker Pro
Maker Pro

Heathkit IP20 Power Supply

mrmodify

Feb 13, 2010
157
Joined
Feb 13, 2010
Messages
157
Evening All, I have a dumb question.

I have a couple of ancient power supplies that I bought years ago. One I have been using for years, I have a new need for the second power supply. I have two power transistors that are open. The parts list list these two transistors as a matched pair.

My question is do they have to be matched pairs and why?

I have inserted a copy of the schematic. The two transistor are located on page 2 Transistors X4 & X5.

Thanks
Happy New Years to all.
 

Attachments

  • ip-20.pdf
    440.6 KB · Views: 299

Bluejets

Oct 5, 2014
6,901
Joined
Oct 5, 2014
Messages
6,901
I think it is because they are operating in parallel and the gains will need to be the same or very close to.
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
RCA 2N2147:
Germanium PNP TO-3
VBEmax 75V; VCEmax 60V
ICmax 5A; PDmax 12W
hFE 10 min at 1A; FT 1 MHz
Suggested equivalents AL100 2N1906

You're unlikely to find anything suitable through the normal electronic component suppliers.You might find something on eBay but you can never be sure of its condition, and in some cases, even that it really is what it says it is. You could also try "antique boutique"-style electronic suppliers for NOS ("new old stock") parts.

The main problem is that it's germanium. Silicon PNPs in TO-3 packages are becoming harder to find too, but Mouser have the Central Semiconductor 2N5880 (http://uk.mouser.com/ProductDetail/...=sGAEpiMZZMshyDBzk1/Wiw99kSkYzPxm8FK5J44PeRU=) in stock for USD 6.63 and you should be able to find one of the old standbys such as the MJ2955.

Ideally the two output transistors should be matched but the two 0.33Ω emitter resistors will help them share the current pretty well.

But that circuit is not designed to use silicon transistors. The biggest difference is the base-emitter forward voltage, which is much lower in germanium transistors. To know whether you can just drop in silicon transistors and adjust the trimpots to compensate, you would have to understand the whole circuit. I've just spent 15 minutes staring at it, and I don't understand it apart from the emitter follower output stage.

There are several reasons - all the voltages are opposite because it's a PNP design (I haven't worked with PNP-centric designs since the 1970s); voltages are all measured relative to the emitters of the output transistors, and there's a toob in there (V1) that I don't know anything about.

But you could try some MJ2955s or similar silicon PNPs in there and see whether you can adjust it for proper operation.

Someone else who understands the circuit would be able to help. Hop, are you there? :)
 

Bluejets

Oct 5, 2014
6,901
Joined
Oct 5, 2014
Messages
6,901
Looks like a diode vacuum tube and associated somehow with the range of the current control and ammeter.
Valves were just going out when I started so in much the same boat as you Kris.
 

davenn

Moderator
Sep 5, 2009
14,254
Joined
Sep 5, 2009
Messages
14,254
Looks like a diode vacuum tube and associated somehow with the range of the current control and ammeter.
No, a diode tube would show a standard bar for the plate as a circuit symbol
also note the black dot inside the symbol .... this indicates gas filled

and it is across the supply rails ... note there is already a diode rectifier just before it

its most likely to be a regulator or visible indicator of some sort
 
Last edited by a moderator:

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
I'm familiar with "normal" valves but that one isn't normal. There's something weird being done with its filament, and the solid dot inside it indicates some kind of gas, IIRC.

Well, I did my due Google diligence and discovered that the 0B2 is a voltage regulator tube. It doesn't even have a heater; it's just been drawn strangely. It only has two terminals. It regulates to typically 108V.

Edit: Dave said most of this already but the quoting in his message was messed up so I didn't see it.
 

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
Silicon transistors will possibly do, even a single one.
I do not understand the point of the thermistor R23 which would have more voltage across it. It may be there to shut down the supply when the temperature goes too high.
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
That's not a 2N2147, it's NTE's "equivalent", NTE121.

From what I've seen, NTE's "equivalents" are often not very appropriate, but in this case, the NTE121 does look suitable, apart from having somewhat lower VCEmax and VCBmax specifications, which aren't significant in this design.

Radio Shack seem to be trying to make a baby with NTE - almost their whole semiconductor range is now NTE parts. And they do have the NTE121 but for a ridiculous price of USD 26.20 each. The link you mentioned has them for slightly less, but still way more than I would pay. The MJ2955 (silicon) should be a few dollars maximum.
 

hevans1944

Hop - AC8NS
Jun 21, 2012
4,878
Joined
Jun 21, 2012
Messages
4,878
...Well, I did my due Google diligence and discovered that the 0B2 is a voltage regulator tube. It doesn't even have a heater; it's just been drawn strangely. It only has two terminals. It regulates to typically 108V. ...
Sorry to jump in so late here... I am sick with something that makes me ache all over... just a cold I hope. Good thing I am fully retired now and can stay in bed, mostly.

The OB2 is a "three-terminal" device, but two of the terminals, connected to the anode, are jumpered inside the envelope. The source supply is connected to one terminal and the load is connected to the other one. This is to prevent the load voltage from soaring to dangerous levels if the toob is removed from its socket with power applied.

I missed the fact that germanium transistors were used. Biasing may be critical, but I haven't printed out the two pages of the PDF schematic so as to paste them side by side, so haven't had a chance to study the circuit in detail. Interesting that Heath used the OB2 to "pre-regulate" the voltage they apply to a zener. I guess that allows most of the power to be dissipated in the 6 W, 120 V light bulb and the OB2 envelope instead of in the zener. Or Heath designers may have had a whole bunch of OB2s they wanted to get rid of...
 
Last edited:

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
Sorry to jump in so late here... I am sick with something that makes me ache all over... just a cold I hope. Good thing I am fully retired now and can stay in bed, mostly.
I'm really sorry to hear that. I hope you feel better soon!
The OB2 is a "three-terminal" device, but two of the terminals, connected to the anode, are jumpered inside the envelope. The source supply is connected to one terminal and the load is connected to the other one. This is to prevent the load voltage from soaring to dangerous levels if the toob is removed from its socket with power applied.
Ah, of course. That makes perfect sense.
I missed the fact that germanium transistors were used. Biasing may be critical, but I haven't printed out the two pages of the PDF schematic so as to paste them side by side, so haven't had a chance to study the circuit in detail.
I've done that; it's attached to this post.
Interesting that Heath used the OB2 to "pre-regulate" the voltage they apply to a zener. I guess that allows most of the power to be dissipated in the 6 V light bulb and the OB2 envelope instead of in the zener. Or Heath designers may have had a whole bunch of OB2s they wanted to get rid of...
Yeah, I thought it was pretty odd too. The really weird thing is that the 0V reference is the negative output terminal on a PNP regulator. That's equivalent to measuring all voltages relative to the positive output terminal of an NPN regulator, which seems pretty silly to me. I guess the design comes from the period when there was still a lot of overlap between toob designs and germanium PNP transistor designs, when no one had really worked out how to deal with the differences properly.
 

Attachments

  • epoint 272122 Heathkit IP-20 power supply schematic.png
    epoint 272122 Heathkit IP-20 power supply schematic.png
    536.5 KB · Views: 919

mrmodify

Feb 13, 2010
157
Joined
Feb 13, 2010
Messages
157
Thanks for all the replies. hevans1944 I hope you get to feeling better. I just got over what it sounds like you have.

I was looking at the NTE121 but they want $24.00 ea here in the US. I'm sorry but i'm a tight wad and want the supply up and running but $50.00 for this power supply to me is a little steep.

I have done some checks with a ohm meter and all else seems to be fine but I don't completely know the condition of the rest of the supply to spend $50.00

I have found some MJ2955 made by Toshiba for $5.00 ea. on Ebay if they are legit.

I am also watching some ECG121 for $27.00 for 2.

Does this group think these will work?

I'm am planning on useing this old power supply to try to do some zinc plating on some old motorcycle parts.
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
I've redrawn enough of the circuit for it to make sense, and yes, I think you'll be fine with two 2N2955s. As duke37 suggested, you could get away with just one, but I guess you'll want it to look as original as possible.

Edit: Or MJ2955s.
 

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
If silicon transistors are used, I would be concerned about the voltage across R23 which is not an ordinary resistor. Some modification to the circuitry may be necessary.

What thinkest thou, Kris
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
D'oh! You're right Trevor. That will be a problem. Thanks for catching that!

Like you, I'm assuming that R23 is an NTC thermistor, attached to the output transistor heatsink. When the output transistors get too hot, its resistance will drop dramatically and it will draw more current and reduce the available base-emitter voltage for the output transistors, reducing the output current and protecting the power supply.

If the output transistors need three times more base-emitter voltage than before, there may never be enough base-emitter voltage to get them conducting properly.

Ideally R23 should be replaced with an NTC with about three times that resistance, i.e. 1500Ω, at 25 °C and a sharp drop in resistance above, say, 70 °C. But I don't know how to find such a device.

You might be able to put a couple of Schottky diodes in series with the existing thermistor I suppose...

I'll have a think about it. If anyone has any other ideas, please suggest them!
 

mrmodify

Feb 13, 2010
157
Joined
Feb 13, 2010
Messages
157
I have no problems going to a single MJ2955. I appreciate all the help. I can go in and check parts, do voltage, amperage, resistance checks. You men are over my head. I appreciate the help in getting this old supply possibly up and running.
 

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
If one output transistor is used,then the emitter resistors R49 and R50 should be connected in parallel. It would be better to use two output transistors to spread the heat better.
I do not fully understand the action of R23 since transistor X3 will try to compensate. Use a couple of Schottky diodes in series with R23 as Kris suggests.

Check X3, if it is shorted, it will overdrive the output transistors.
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
Don't worry, I haven't forgotten you! I'm trying to draw it out in a way that I can understand, so I can figure out the best way to handle the R23 problem.

If you have, or can find, any other documentation on the IP-20, that would be very helpful.

It would be better to use two output transistors to spread the heat better.
Yes, I agree with that.
I do not fully understand the action of R23 since transistor X3 will try to compensate.
Yes it will, but I think R9 and X3's current gain will limit the amount of current it can supply. I'm assuming R23's resistance drops pretty steeply once the temperature gets high, and this would swamp any variation in the amount of drive coming from X3 and force the output transistors to cut back at some reasonably well-controlled temperature.

It might be simpler to add a circuit to monitor R23's resistance and when it gets too low, kill the whole power supply through the latching shutdown relay.

mrmodify, how important is it for the unit to look original inside? And for the overtemperature protection to operate as it did originally?

BTW if you haven't seen it, here's a brief introductory video for the Heathkit IP-20 by Jeff Tranter (that name rings a bell but I can't place it).

 

mrmodify

Feb 13, 2010
157
Joined
Feb 13, 2010
Messages
157
Evening all,

I just purchased the MJ2955's.

Check X3, if it is shorted, it will overdrive the output transistors.

Tested X3 and it checks out with a transistor checker that I have. I do think I have a issue with X1. It checks good but I need to research the number. It does not match the parts list.

Don't worry, I haven't forgotten you! I'm trying to draw it out in a way that I can understand, so I can figure out the best way to handle the R23 problem.

I appreciate you taking the time. I have other projects going on at the same time. time is not a issue.

If you have, or can find, any other documentation on the IP-20, that would be very helpful.

I tried to attach the owner manual. actually its a building manual but it is to large to upload. It is in PDF and I'm not wood with Adobe. But I will see what i cn do.

mrmodify, how important is it for the unit to look original inside? And for the overtemperature protection to operate as it did originally?

Its not important to me to keep the insides original, I would like not to destroy it but if I do I wont be upset. It will go back into my parts bin and become a spare to the other one I use. I am going to use it for platting if this works out. I just didn't want to buy a power supply to plate parts.

I very much appreciate the time and mental thinking this project. I am good mechanically and below fair on electronics with out you all this project would probably be dead.

Thanks John
 
Top