• Please review our updated Terms and Rules here

11/34 Power Supply Connector Question

SteveG

Experienced Member
Joined
Sep 30, 2020
Messages
151
Location
Millmeece Near Stafford UK
I am stuck disconnecting the Power Supply from the body of my 11/34 !

As well as the usual plastic side locks there is a metal clip sticking out of the side of the connector (see photo).

It would help if anyone knows:-

1) If this clip could be preventing me from releasing the cable?

2) What is it’s purpose of this clip as it is not keen to move in and out
 

Attachments

  • 97042581-563E-4231-82EF-9426C803C9C7.jpeg
    97042581-563E-4231-82EF-9426C803C9C7.jpeg
    125.7 KB · Views: 1
IIRC that plastic tab is used when the connector is panel mounted. It locks the connector into the panel.

joe
 
Thanks Joe,

So I can ignore this clip as it will not be holding anything together !

The large 11/34A - HJ 7amp Power Supply is now liberated.

Thanks
 
Last edited:
I am stuck disconnecting the Power Supply from the body of my 11/34 !

As well as the usual plastic side locks there is a metal clip sticking out of the side of the connector (see photo).

It would help if anyone knows:-

1) If this clip could be preventing me from releasing the cable?

2) What is it’s purpose of this clip as it is not keen to move in and out

As noted earlier that clip serves another unrelated purpose. The "not keen to move" part is probably a combination of factors:

1) You need to squeeze the retention clips at the ends so that those tabs can be extracted through their surrounding rectangular "housing", and

2) There is probably some contact oxidation to be overcome.

Suggest freeing one end with squeeze-and-rock, then the other end, then keep rocking. Don't pull on the wires, and do ensure that the support for the other half of the connector is firm (e.g., you're not over-flexxing a PCB). You probably will need to apply pliers -- suitably sized, positioned and perhaps padded with a piece of firm rubber
 
Hello Pbirkel,

I have done what you recommended and the plug EVENTUALLY came apart.

I have not had time to establish if any damage has occurred yet but so far so good.

What a WIERD power supply. Four separate units and one main large transformer (see photo).

I suspect it will be best to test each unit separately. Can anyone point me to the inputs and outputs required for each of the four Power Supply units.
 

Attachments

  • 5D305CB2-131A-47E6-9E83-50F0F434AF9D.jpeg
    5D305CB2-131A-47E6-9E83-50F0F434AF9D.jpeg
    138.7 KB · Views: 1
  • 4495F083-E991-49F0-B187-A1A8803E532F.jpeg
    4495F083-E991-49F0-B187-A1A8803E532F.jpeg
    157.4 KB · Views: 1
>>> What a WIERD power supply. Four separate units and one main large transformer.

Standard fare for DEC...

I think the information you need is all in here: http://www.bitsavers.org/pdf/dec/pdp...Vol2_Sep76.pdf. Towards the end of the manual.

The first thing to check out is the transformer itself. Each secondary winding (well, the ones feeding the regulators) should output approximately 28V ac (if memory serves me correctly). Of course, before powering up, make sure that the system is wired for the correct voltage (240 or 110 Volts), all of the fuses are of the correct rating and that you have at least performed some earth bond and insulation tests - or checked that your life insurance policy is up to date!

You can generally power each separate regulator from a bench power supply - much safer that way!

Dave
 
I plan to test the four regulators on the bench first.

Fortunately they are well label as can be seen on the photo below.

I don’t have a 20 - 30 V AC bench power source so I was thinking of using a Variac to reduce the AC from 240 to 20 - 30 (Is that safe?).

I am planing just to check the Regulated DC voltage with a meter. This assumes I don’t need to apply a load to check the regulator operation. (Is this correct?).

I guess you can see from my questions that my experience is software and not hardware.
 

Attachments

  • A5A0D840-A06C-462A-A633-CDB1D7D3BDBE.jpeg
    A5A0D840-A06C-462A-A633-CDB1D7D3BDBE.jpeg
    158.5 KB · Views: 2
Nice project! I have a BA11K expansion box with a bunch of memory in it which I have to check/restore. Seems to be the same as your 11/34 chassis.
 
I don't have a 20 - 30 V AC bench power source so I was thinking of using a Variac to reduce the AC from 240 to 20 - 30 (Is that safe?)

It is absolutely Not safe!
A Variac provides no isolation, so your regulators could be live relative to Earth (ie: You and your oscilloscope) if the live and neutral are swapped. Even if neutral goes to the "earthy" side of the variac, a fault condition could still make it dangerously above earth potential. Do not do this!

A better alternative is find some heavy duty transformer of almost any primary and secondary voltages, and figure out a way to get 28VAC from one side of it (primary or secondary), while feeding the other side of it from the variac set to the appropriate voltage.
Maybe even a rewound microwave oven transformer would do the job for short term testing. Especially if you reduce core saturation by you winding it for say, 36VAC output, then feed it from the variac with 80% or less of it's normal mains input voltage. (and remove the magnetic shunts as well). I believe that MOT secondaries work at about 1v/turn, so saw off the high voltage secondary, and wind 36 heavy gauge turns in its place. Any errors are easily adjusted out by the variac.

Remember also that variacs are not capable of their full rated power when operating at the lower end of the dial.
 
I would second the last post!

Buy a bench power supply. It will come in useful for all sorts of projects. I have a 2 channel variable output power supply. The unit has inbuilt voltage and current meters for each channel and the ability to set a current limit.

I’ll see if we are junking any from work for you...

All switch mode power supplies should have s load of some sort connected to them. A trip to Halfords to pick up various bulbs... Some motorcycle lamps are 6V, so ideal for +5V regulators. Use side lights, brake lights, dip and main beam lights in various parallel/series combinations to soak test / stress test the power supplies before applying power to your delicate electronics.

You could also obtain some BIG power resistors, heat sink and a bucket of water!

Dave
 
Light bulbs make a simple dummy load, but their high positive temperature coefficient can also make them tricky. When cold their resistance can be 15 times lower than when hot, which under some circumstances can make the PSU oscillate wildly into and out of current limit as the bulb heats and cools. So getting a reasonable load when hot (and stable), means going through a much higher load at startup, which could cause problems.

By comparison, resistors have much lower PTC. I tested my H777 and H740 PSUs using some power resistor banks in an oil tank, (but had some serious problems with using them on the H740s - and a variac).
These threads I started may be of interest:-
https://www.vcfed.org/forum/forum/genres/dec/71298-oil-filled-dummy-load-for-h777-psu
https://www.vcfed.org/forum/forum/genres/dec/1220891-dummy-load-for-dec-pdp-computer-power-supplies

(BTW: search for "rewind a microwave oven transformer" on YouTube for all you need to know on that topic)
 
You don't need an AC power supply for these regulator modules as they can just as easily be powered from DC. The input goes immediately through a bridge rectifier so you can connect up a DC bench supply with any polarity. As for a load I also use several low value resistors in parallel; the 50 watt metal body type that I clamp to a heat sink (usually a large piece of scrap metal). Some power supplies can be damaged by running them without a load so it's best to always use one.
 
Thanks for all the advice !

I have a variety of DC power supplies so I think I will start with them.

I have a handy little PC power supply tester (see photo) which has worked well for the smaller museum computers.

I wonder how this little unit applies a load in its testing cycle. It does not give the impression that it will dissipate much heat!

NB. I have discounted trying to use this to test the larger vintage DEC supplies.

PS. A visual inspection of the Regulators makes me think that I should probably replace the following capacitor (see photo).

I may run a test anyway without replacing it to enable a “sweepstake” on the outcome.
 

Attachments

  • C138BB63-4B4E-4390-A8AB-C6A6FCA04CE7.jpeg
    C138BB63-4B4E-4390-A8AB-C6A6FCA04CE7.jpeg
    179.7 KB · Views: 1
  • A9EE72E7-ADDA-4C72-84B5-E2C180496F54.jpeg
    A9EE72E7-ADDA-4C72-84B5-E2C180496F54.jpeg
    143.1 KB · Views: 1
Last edited:
You are correct storage conditions were not ideal. However other computers including some DEC VAX’s stored alongside are now working after various amounts of remedial work.
 
Hello Pbirkel,
What a WIERD power supply. Four separate units and one main large transformer (see photo).

I suspect it will be best to test each unit separately. Can anyone point me to the inputs and outputs required for each of the four Power Supply units.
I was just in several of these chassis this past weekend (11/70 MJ11, 11/34 and 11/04) at the LSSM. These were a very popular supply at DEC and you should consider yourself lucky that you don't have something like the 11/44's H7140. Which I was (unfortunately) also in.

I see from your subsequent post that you have 4 power supplies. The H754 indicates that this box has core memory (or at least did at one point). Two supplies being H7441 instead of H744 means that this is likely an 11/34A and not an 11/34.

Anyway, you can just measure the output of these supplies un-loaded. They may read a bit higher than when loaded, but at least you can confirm that the voltages are reasonable. I just pull the CPU and any peripheral boards and test the whole thing in-chassis. That way you don't have to deal with the "pulling teeth" of getting those supplies out unless one of them reads improperly.

BTW, the MJ11 had not been powered up since Field Service converted the 11/70 (at their expense) to semiconductor memory decades ago. It was fine. The 11/04 was also fine and hadn't been powered on in some unknown-by-me amount of time. The 11/34 had been working, blew a H7441 which was replaced in the past, but the system still didn't work after the replacement. It turns out the H744 that was put in was reading 4.5V when loaded and couldn't be adjusted past 4.85V (that was where I ran out of trimpot). A H7441 from spares got that system working fine again.
 
Hi Terry, I'd be interested in hearing more about your H7140 troubleshooting, I have a one in an 11/24 that I'm unable to get working. It is supplying the 35V fan voltage but no other voltages are present.
 
Back
Top