UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Specific Vintage Equipment > Vintage Test Gear and Workshop Equipment

Notices

Vintage Test Gear and Workshop Equipment For discussions about vintage test gear and workshop equipment such as coil winders.

Closed Thread
 
Thread Tools
Old 7th Feb 2017, 3:12 pm   #1
MrBungle
Dekatron
 
MrBungle's Avatar
 
Join Date: Jun 2016
Location: London, UK.
Posts: 3,687
Default Heathkit IP-2718 power supply annoying failure - any ideas?

Just had an interesting hour of debugging. My Heathkit IP-2718 just decided to incinerate the breadboard I was working on. Was designing a discrete differential amplifier to work across an 18v supply differential (2x 9v batteries). Next thing I know, smoke was pouring out of one of the clamped 2n3904's making up the pair. Spent about 10 minutes working out what had gone on, then I noticed that the thing was getting rather warm.

Measured the output voltage (supposed to be 9v) and was surprised to see 37 volts across each of the supplies bringing the total destruction up to a nice 72 volts! I had the metering set on the 5v rail so I didn't see the current or voltage jump.

Had a poke around with a meter and found both the pass transistor drivers for each half of the supply, which were MPSA42's had gone C-E short! No idea what has caused that. Especially both which is weird!

I actually had a couple of MPSA42's floating around so I soldered one in and it's absolutely fine now on that rail. I am currently replacing the other one.

No idea what caused it. Wondering if there was a surge or something that knocked the rail voltage up or something odd.

Any ideas or hypothesis' would be appreciated on this one as I can't see a failure mode that would shoot both. They are running as a 9-0-9 supply for reference.

Schematic is here: http://www.tauntek.com/Heathkit-IP-2718.pdf (last two pages)
MrBungle is offline  
Old 8th Feb 2017, 8:49 pm   #2
The Philpott
Dekatron
 
The Philpott's Avatar
 
Join Date: Aug 2016
Location: Colchester, Essex, UK.
Posts: 4,081
Default Re: Heathkit IP-2718 power supply annoying failure - any ideas?

Is there a capacitor in there that has died?
The Philpott is offline  
Old 9th Feb 2017, 12:42 am   #3
threeseven
Octode
 
threeseven's Avatar
 
Join Date: May 2010
Location: Worcestershire, UK.
Posts: 1,880
Default Re: Heathkit IP-2718 power supply annoying failure - any ideas?

Very strange, considering that each half is independent apart from the transformer! Even a surge should have been compensated. Load current would be low, so the pass transistors were hardly being taxed. Very odd.
__________________
BVWS member
threeseven is offline  
Old 9th Feb 2017, 8:01 am   #4
MrBungle
Dekatron
 
MrBungle's Avatar
 
Join Date: Jun 2016
Location: London, UK.
Posts: 3,687
Default Re: Heathkit IP-2718 power supply annoying failure - any ideas?

Exactly my problem. This is a strange one. It is behaving absolutely fine since the event though so I may chalk this one up to gremlins.

Have checked all capacitors and all are fine. In fact they were all replaced with high quality BC/Vishay long life ones about 9 months ago.
MrBungle is offline  
Old 9th Feb 2017, 8:17 am   #5
G4_Pete
Hexode
 
Join Date: Feb 2014
Location: Southampton, Hampshire, UK.
Posts: 419
Default Re: Heathkit IP-2718 power supply annoying failure - any ideas?

Here is an Hypothesis to check "If one pass transistor fails and the A channel is connected to the B channel via an external circuit then the B channel would also be damaged”
The null Hypothesis being that no damage would occur in the above scenario.

I see there is no diodes to protect back feed if I read the diagram right.

Pete
G4_Pete is offline  
Old 9th Feb 2017, 8:37 am   #6
G4_Pete
Hexode
 
Join Date: Feb 2014
Location: Southampton, Hampshire, UK.
Posts: 419
Default Re: Heathkit IP-2718 power supply annoying failure - any ideas?

Hi, Just thinking a bit more was the PSU strapped for +/- on the psu or via the breadboard as just thinking if a breadboard link could have caused the A/B linking?

Just thoughts -I may be barking up the wrong tree completly!!

Pete
G4_Pete is offline  
Old 9th Feb 2017, 11:46 am   #7
MrBungle
Dekatron
 
MrBungle's Avatar
 
Join Date: Jun 2016
Location: London, UK.
Posts: 3,687
Default Re: Heathkit IP-2718 power supply annoying failure - any ideas?

There are indeed no protection diodes there at all. There aren't any inside it either. That's actually an interesting hypothesis. Usually when designing them you put one across the output and one across the regulator to stop reverse overvoltages blowing the thing out.

The supplies were linked as follows:

Click image for larger version

Name:	Untitled.png
Views:	76
Size:	3.5 KB
ID:	137308

I have shorted the outputs on these supplies plenty of times and they switch into CC mode and limit to around 500mA and get a little warm but that is about it.

I've gone back to my nice TTi supply for the moment though as I don't trust this one until I've worked out why it went nuts.

Last edited by MrBungle; 9th Feb 2017 at 12:01 pm.
MrBungle is offline  
Old 10th Feb 2017, 10:57 am   #8
dave cox
Nonode
 
dave cox's Avatar
 
Join Date: Jan 2009
Location: Bristol, UK.
Posts: 2,059
Default Re: Heathkit IP-2718 power supply annoying failure - any ideas?

MPSA42 seems like a bit of a light weight to me !

At max OP current (and an output device with min dc gain ~ 20) when outputting ~ 0V would put 30+ volts across the driver at 25mA (plus whatever the BE resistor sucks up). Probably the gain isn't as low as 20 so it would 'most likely' be OK

Another factor is once one has failed (with its OP turned full on) the load may try to pull the other output below ground ...

dc
dave cox is online now  
Old 10th Feb 2017, 12:04 pm   #9
MrBungle
Dekatron
 
MrBungle's Avatar
 
Join Date: Jun 2016
Location: London, UK.
Posts: 3,687
Default Re: Heathkit IP-2718 power supply annoying failure - any ideas?

The MPSA42 is a typical Heathkit design i.e. "what have we got in inventory that'll just about work". I have some serious problems with the power supply design on their IM-2410 counter. Not enough juice? Stick a big carbon comp resistor across the regulator to shunt some current around it and use it to just regulate that. Gets a bit hot but clearly cheaper than a TO-3 regulator at the time. Anyway, I digress...

That's a good point with respect to the cascading failure possibility due to the output being pulled. I suspect they didn't go pop at precisely the same moment. I've drawn up some changes to the power supply to make things a little nicer. I'm going to swap the MPSA42's out for a couple of 2n2219A's as they do indeed get warm. Voltages are fine on the 2n2219A and it can shift between 0.8 and 3.0W and 800mA.

Also going to add protection diodes across the regulator transistor and across the output.

I'll give Heathkit their due though; the engineering is pretty good generally apart from a few notable turds.
MrBungle is offline  
Old 10th Feb 2017, 3:18 pm   #10
MrBungle
Dekatron
 
MrBungle's Avatar
 
Join Date: Jun 2016
Location: London, UK.
Posts: 3,687
Default Re: Heathkit IP-2718 power supply annoying failure - any ideas?

I just spent the last two hours taking this to bits and I just found the problem. The B supply is still failed to 37v after replacing the transistor. I didn't notice this at first as I was half asleep by the looks. Turns out the B supply reference buffer transistor is knackered meaning no regulation is taking place. I assume this did pull the A supply down and a cascading failure blew up both of the MPSA42's as this kills the current limiting completely as well. The failed transistor was an MPSA20. Vce is a little close at 40v and it has 34 across it which may have contributed to the failure. Also the soldering around there suggests it has been replaced by a previous owner before as well! I replaced both Q101/Q201 with modern 2N3904 and both Q203/Q103 with 2N2219As and we're in business again. 2N3904's will handle a good 60v across Vce without failure from experience.

Fingers crossed!
MrBungle is offline  
Old 10th Feb 2017, 5:43 pm   #11
G3VFB_Anton
Hexode
 
G3VFB_Anton's Avatar
 
Join Date: May 2016
Location: Taunton, Somerset, UK.
Posts: 348
Default Re: Heathkit IP-2718 power supply annoying failure - any ideas?

Slightly off topic but nevertheless a comment on Heathkit PSUs mainly the IP-20 built 1962-67. Recently discovered mine and recall having to fix it a few times with the then not very reliable power transistors often going o/c. Not yet opened or tried it to see if the OC35 put in to replace the two 2N2147s is still functional. A very useful piece of kit
__________________
Anton

Golf Three Very Fine Business
Member of RSARS and PARC
G3VFB_Anton is offline  
Old 11th Feb 2017, 4:27 pm   #12
wd40addict
Octode
 
wd40addict's Avatar
 
Join Date: May 2009
Location: Stevenage, Herts. UK.
Posts: 1,515
Default Re: Heathkit IP-2718 power supply annoying failure - any ideas?

I 'siliconised' my IP-20u.

When I got it it wouldn't go above approx 25v. Turned out a previous owner had a cascade fault and had changed most of the transistors, however he'd used something with too low a Vce rating to replace the 2N398.

Rather than trying to source obscure Germaniums I changed them all except the current limit pass transistor which was OK.

In the course of this work I found out what had probably caused the original fault - the heatsink insulation had not been installed correctly and it had probably shorted to chassis.

Note also that the original IP-20 circuit has no real protection for reverse bias on the 2N398 EB junction. This was corrected later by fitting a diode and re-arranging the position of the cutout relay. The UK version has this mod as does the later IP-27 which basically uses the same circuit.
wd40addict is offline  
Old 11th Feb 2017, 5:35 pm   #13
MrBungle
Dekatron
 
MrBungle's Avatar
 
Join Date: Jun 2016
Location: London, UK.
Posts: 3,687
Default Re: Heathkit IP-2718 power supply annoying failure - any ideas?

Ironically, my heathkit counter had a badly installed insulator on the power transistor as well. Replaced it and all is good.

YMMV with Heathkit. Sometimes it's really well built, sometimes it's awful!
MrBungle is offline  
Old 21st Feb 2017, 11:37 am   #14
radiotechnician
Pentode
 
Join Date: Jan 2013
Location: Powell River, British Columbia, Canada.
Posts: 217
Default Re: Heathkit IP-2718 power supply annoying failure - any ideas?

I own 6 of the IP2718 power supplies. I built one when they came out,
and bought the rest in a lot from a college.

I teach electrics/electronics on contract and furnish my own lab
equipment.



The last time one of the 2718s went up the spout it was an
open resistor and a MPSA 42. I pinched a spare out of an
old Heathkit desk calculator kit.

Since then Electronic Goldmine had MPSA42s for pennies,
and I bought a slew because they are useful for neon drivers.
__________________
Steve Dow
VE7ASO
radiotechnician is offline  
Old 21st Feb 2017, 11:55 am   #15
MrBungle
Dekatron
 
MrBungle's Avatar
 
Join Date: Jun 2016
Location: London, UK.
Posts: 3,687
Default Re: Heathkit IP-2718 power supply annoying failure - any ideas?

They're nice little power supplies so I can see why you use them for teaching!

Got my MPSA42's from here: https://www.bitsbox.co.uk/index.php?...j1384ivb105565

They were purchased to drive CRT deflection in a diff amp config. Require a little heatsink for that though!
MrBungle is offline  
Old 20th Apr 2017, 10:33 pm   #16
MrBungle
Dekatron
 
MrBungle's Avatar
 
Join Date: Jun 2016
Location: London, UK.
Posts: 3,687
Default Re: Heathkit IP-2718 power supply annoying failure - any ideas?

A minor update on this thread.

It blew up again. Fortunately it was connected to a zener regulated VFO on my breadboard. No smoke this time

I have however found the problem. The original constructor had cocked up with some wire routing and soldering and had melted through a wire which was contacting the leg on the MPSA42 that died. As always it's the silly problems you don't notice! I have replaced the entire wire and inspected all of the others and this was the only problem.

Interestingly, I hit ebay as I regularly do looking for Heathkit stuff (it appears I'm collecting it) and found an identical and incredibly clean looking supply purely by chance for £25 so I grabbed that.

I love these power supplies so much I need a lifetime supply of them
MrBungle is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 11:28 pm.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.