looking (again) for a PSU guru.

[iainmaoileoin]

Some of my 11/84 PSUs are not working (working well).

I am trying to find somebody with any skill or knowledge in debugging this hardware - H7202B.

At the very least if somebody could explain to me what signals need to be held low to get the PSU to (try to) start
that would be great.

J5
pin 14 Mod En - leave floating?
pin 17 Ignore
pin 12 hold low - NOT(PWR. REQ.)
pin 11 hold low - NOT(SWTCN)
pin 10 leave floating - NOT(STBY)

I am even less sure about pin 19 - NOT(AIR FLT)


Any clues - of is there a better group for hardware issues?

[iainmaoileoin]

Replying to my own question - in part
section4.9.jpg

It would look like I only need to assert PWR REQ (take it low)

On a better scan SWTCN is actually SHTDN - shutdown - so I leave that high.

I now dont understand the "return" (ground?) that each power supply has.
If I am testing "out of circuit" do I connect all these returns together?

Thanks in advance

[Qbus]

can you post a link to a schematic? too lazy to look up on my own. do you have a scope and a load to put on the power supply when its removed from the hardware? Have played around with H720 first generation switching power supplies before and they can be a bear. biggest issue with those PWM switching supplies is all the protection circuits and getting everything to let it run.

[Bitly]

AIR FLT should be 'Air Fault'. On the 780's, there are vanes connected to microswitches that stick into the airflow under the card cage. If the fans aren't running (or blocked), the switches prevent powerup. Holding one of these switches in your rack should indicate the polarity required.

CW

[iainmaoileoin]

can you post a link to a schematic? too lazy to look up on my own. do you have a scope and a load to put on the power supply when its removed from the hardware? Have played around with H720 first generation switching power supplies before and they can be a bear. biggest issue with those PWM switching supplies is all the protection circuits and getting everything to let it run.

Of course. MP01270_11730_Engineering_Drawings_Apr82
About page 50 to about 90. I can cut out the relevant bit if the 300 or so pages are too big to load.

I have scopes, variac, ESR, AVO (that old), load resistors, login analysers - old and big. Some soldering irons (old and big), some burnt fingers I still cant aim an iron, some solder and hold a component without getting myself! I have some skill on TTL/CMOS, I can do Carver and Mead 1980s VLSI, very limited knowledge past BC108/109s, I hate 555s and I could never master op-amps (I always had to steal somebody elses design!).

Right now I am struggling to get a known working beast working "outside" the cabinet. I am just not driving the right control signal(s). That should be the easy bit - since it is all CMOS. Damned if I can get the working PSU to work. Perhaps not enough load? Embarrassing!

Also I am stuck to know what to do with the "return" signals for each of the separate PSUs.

Any help/input will be welcome.

The plan is to get some of the kit into my museum (see) - big pictures/slow.

One of my PSUs starts up 1 or 2 in 10 power-ons - provides all voltages and allows me to boot to the console. But often the 5volt rate drops out - a few seconds after starts - NEVER once it runs. Looks like the device is going into standby. A know good PSU in the same chassis runs fine.

The 2nd PSU just wont fire up at all. I am not keen to connect to a good chassis - I have not idea of the origin of that supply.
Caps, resistors, PCB all visualled, joints resoldered if anything looked dry.

Iain

[Qbus]

Man that 7200 series is a bear! Looks like that’s what you need for 5 volts at 60 amps though. When it was in the system did the memory and communications power supplies work? Or did all fail when it went into fault?
Be mindful of the primary side of the power supply is Hot to ground and such is isolated from the system or frame ground, but imagine you already know this.
Going to assume that you also know that the +150 and -150 sources are good in respect to the neutral buss.
Would be good if R1, R2 or R5 and R6 were week or open but don’t think it will be that simple. Will look at the interlocks and try to get an idea of what needs to be connected for the control to allow the power supply to come up and run but that may take some time.
There is also a +14 and – 14 volt power source that’s developed by T4 with its own oscillator and switching supply that powers all the power supply logic and control and if you don’t have +14 to ground C32/34 and -14 to ground at C33/35 nothing is going to happen anywhere else.
The +14/-14 volt power should be there the entire time the +150/-150 volt bus is up, regardless of interlocks or connections. The incoming AC power is sampled from the hot side of the AC line by the control for the T4 power supply regardless of what the rest of the power supply is up to.
Whenever the system is connected to power the T4 power supply is running.
Again, be careful being everything on the primary side of the T4 power supply is at line potential. Have to wonder what you have there for AC power? Assume its 220 Volts 50 cycle. Are both sides of your AC line isolated from ground or is one side neutral (grounded) and the other at 220volts? US systems are one side at neutral or ground and one side at 120 Volts with a US 220 volt service being each side together makes 240 volts with each leg being 120 volts to neutral.
The power supply has both the nigh leg, labeled PHASE and the low leg labeled Neutral isolated from the frame and safety grounds so it won’t matter just curious on my part.

[iainmaoileoin]

Man that 7200 series is a bear! Looks like that’s what you need for 5 volts at 60 amps though. When it was in the system did the memory and communications power supplies work? Or did all fail when it went into fault?

One supply has a 5V problem only. The other wont produce any volts. I dont know how it failed. I do know - no smoke - no bang!

I am 1/2 tempted to replace the multi-psus with a couple of new switch-modes, but that would be cheating!

Be mindful of the primary side of the power supply is Hot to ground and such is isolated from the system or frame ground, but imagine you already know this.

One of the reasons I just hate working on switchmode - My eyes and fingers dont work as well as they used to. I lose more bits by the occasional "oops" than I ever fix!

Going to assume that you also know that the +150 and -150 sources are good in respect to the neutral buss.
Would be good if R1, R2 or R5 and R6 were week or open but don’t think it will be that simple. Will look at the interlocks and try to get an idea of what needs to be connected for the control to allow the power supply to come up and run but that may take some time.
There is also a +14 and – 14 volt power source that’s developed by T4 with its own oscillator and switching supply that powers all the power supply logic and control and if you don’t have +14 to ground C32/34 and -14 to ground at C33/35 nothing is going to happen anywhere else.

The +14/-14 volt power should be there the entire time the +150/-150 volt bus is up, regardless of interlocks or connections. The incoming AC power is sampled from the hot side of the AC line by the control for the T4 power supply regardless of what the rest of the power supply is up to.
Whenever the system is connected to power the T4 power supply is running.

I follow your logic - I will check that out.

Again, be careful being everything on the primary side of the T4 power supply is at line potential. Have to wonder what you have there for AC power? Assume its 220 Volts 50 cycle. Are both sides of your AC line isolated from ground or is one side neutral (grounded) and the other at 220volts? US systems are one side at neutral or ground and one side at 120 Volts with a US 220 volt service being each side together makes 240 volts with each leg being 120 volts to neutral.
The power supply has both the nigh leg, labeled PHASE and the low leg labeled Neutral isolated from the frame and safety grounds so it won’t matter just curious on my part.

It used to be 240/250 volt single-phase in the UK. Some years ago it was harmonised with Europe and become 220. I still call it 240volt.
We run LIVE and NEUTRAL to things. LIVE is fused and RCB checked (for leakage), Earth is a separate line that visits most metal-work/devices etc. In the case of double insulated kit the earth does not pass through the power supply (Power supplies for many laptops have a "floating ground" not connected to earth. In the UK such kit is sometimes called "class 2" and is identified with a double square logo on the product.

Generally speaking all racks (and panels) and the chassis of our kit are earth bonded. Usually, in data-centres the earth is NOT the local plug earth but runs directly from the distribution board (say fuse box).

[g4ugm]

Iain,
Its actually 230v. They didn't change the voltage at the wall. In most places its still 240v. Typical politicians they changed to tolerance so its now 230V +10%/−6% which is between 216 to 253v. They have talked about lowering it to save energy but with so many switched mode devices that simply adjust the load the actual savings are minimal..
Dave

[Qbus]

Ok, still think the two things that you have to determine are if you have the +150 and -150 volt sources that can be seen at J1 pins 9 for +150 and 4 for -150 you can also see this at the two big capacitors C1 and C2, the junction of the two capacitors is at AC Neutral or the return ground for the power supply but remember that’s not the chassis ground. And the two far ends of the capacitors will be +150 and -150
That’s the source voltage for feeding all the switching supplies and is always there as long as primary power is connected.
That voltage feeds a small power supply that provides +14 and -14 volts that’s used to control and drive all the parts of the power supplies and you can see that voltage on J2 or J3 on pin 6 for +14 and pin 7 for -14.
The +/- 14 volt power supply is derived from the +/- 150 volt source and just like that it’s always there as long as primary power is connected.
It’s disconcerting to think that regardless of if the system is on or off a lot of the power supply is up and running. A good argument for removing the power cable when not in service or at least using one of the DEC 861 power control centers for feeding everything.
My Qbus and Unibus systems both have their own racks each with an 861 in the bottom of the rack that’s controlled thru the power switches on the CPU

[iainmaoileoin]

Qbus: Ta
I have a few projects on the go. My benches are full of the 11/70 front panel (et al) and I hope to have that beast in a box by Thursday/Friday.
Then I will get these PSUs back out. Your clues are very welcome.

Dave: Thanks for the update - I live in a rural area at the end of long set of wooden poles - with a transformer bolted at the top. My volts range from 0 to 250! It is closer to 250 most of the time. When my next door neighbour (about 1/2 mile down the road) brings on his wood lathe and or surfacer, I get a real dip in volts in the house. He has a single-phase to 3phase 'fooler' for his professional kit but it is hefty at startup of the motors.