Latest acquisition

[insurgent]

Having no DC bench power supply at all I've been scouring EBay for deals on older BK Precision models. I have this strange obsession for a triple or quad output and the prices for those type of modern units were just too far out of my reach at the moment. I had set my sights on a new, old stock BK Precision 1650 (2 x 0-25v @ 0.5a, 1 x 5v @ 5a) as my starter supply but the listing had ended. The fellow said he would re-list it in a week or so but fate intervened and I stumbled across another supply the day before he re-listed:


Pardon the ghetto desktop random-bits-of-wood stand that it's on at the moment; in progress house renovations (I'm sure Dave can sympathize) have left me with about 8 square feed of space for my computer/electronics lair.
It's a Sorensen XTQ 20-3 with remote analog programming cards. The 4 supplies can be independent, series or parallel (though you manually have to set this up with wiring) and support CV/CC and current limiting. It also has "sense" connections that are used to compensate for losses in the output power leads. Unfortunately it doesn't have the option installed for the master/slave tracking but I'm thinking down the line I can use a microcontroller to accomplish this via the remote programming interface. The voltage pot is a 10 turn pot and the current pot is a single turn. Apparently the 10 turn current pot was another "option" that isn't on this unit.

The "Sorensen, A Raytheon Company" on the back places it in the late 80's early 90's but the specs that I could find were quite good and the unit seemed fairly operational. The line/load current regulation specs seem to be better on this old unit than the new $500 units (and 4 x $500 is a buttload of bills)! I figured that even if only 2 of the 4 PS's were operational that it would still be a decent deal. I got the unit for $150 shipped though I had to buy a manual for an additional $25 from someone else as I couldn't find it on the net. Fortunately the manuals from back then actually have schematics and theory of operation descriptions! Apparently all calibration is done by replacing specific socketed components of identical value.
The display reading on units 2 & 3 constantly "wander" (e.g. randomly fluctuating from 19.7V to 21.9V when set at 20V) and there's a nice transformer hum (a tad louder than the Rigol's fan) though the outputs are rock solid. All units output 0-20V and 0-3A. I can't tell how accurate the voltage/displays are until the Agilent meter shows up next week  ( thanks to Russel for the heads-up! ) but on my Radio Shack meter (4k count, 0.3% DCV accuracy) shows the output rock solid and the meters off a little (e.g. Setting all units to 10V according to the meter read: 10.1, 8.9-10.8 (wanders), 8.8-10.9 (wanders), 10.0)

I've attempted to measure the ripple on the Rigol and, assuming the settings I used are correct for measuring this, these are the results (probes 1x, y = 2mV, x = 100ms):
Unit 1 & 4

Unit 2 & 3

I used the spring clip on the probes instead of the alligator clips attached as seen in the picture above. Doesn't seem to me to be damned good for 20+ year old supplies. The manual spec is < 1mVrms.

If anyone can suggest other tests to run on the units, let me know (and explain how!). I will be able to better determine the voltages when the Agilent meter arrives this coming week.
I'll post some internal teardown pictures if anyone is interested.

Take care,

John

[saturation]

Nice!  Congratulations.  Will have more later.  Just got my Agilent DMM from the Grainger.com deal, its sweet.  If this is the DMM you're getting, I'm sure you'll be happy.

Its old stock, manufacture date is late 11/2007!, calibration date is 3/2009, so its been sitting unsold for ~ 2 years, and over 3 years old.

[dimlow]

because of the age of this it may well be worth getting inside and checking the caps for ESR and leaks. Could save a head ache later.

[insurgent]

because of the age of this it may well be worth getting inside and checking the caps for ESR and leaks. Could save a head ache later.

I'll tell the wife that you said I *had* to get another test instrument

Looking around, I could test ESR in-circuit using a 555 circuit and o-scope (or drop $90 on a "blue" meter which people seem to like). To my knowledge it's not possible to test leakage in-circuit, is it?

[dimlow]

By leakage I meant gunk on the PCB and the smell of, well err, cat pee best describes it! But you dont really get the cat pee stink until you try to remove the caps.

For ESR i use the ESR-micro V4.0 from a Russian guy, it works well and i think was only $60 US.

[tekfan]

Sorensen - one of the best power supply manufacturers I know of.
150 USD - I am very very envious. This is one of the reasons why I would move to the states.

If you really want to do noise measurments with an oscilloscope it's rather difficult. With normal single ended scope probes you pick up all sorts of nasty things such as noise, hum, ground currents. So it's no surprise that your scope showed that kind of noise level. You can minimise this by using an isolation transformer and letting the power supply chassis float and be connected to the scope ground and the scope to mains earth. That way you don't create ground loops. Or you could use a scope with a differential input. Tektronix 7000 series are perfect scopes for the job in this case. You can get differential amplifiers that measure to as low as 10 microvolts per division. In the US you can get them for next to nothing. Make sure the CRT isn't broken (very very rare but worth taking a look) and the rest is repairable. The scopes are very big compared to the Rigol but worth every penny.

I think putting a multimeter on the lowest AC range will give you the most accurate reading if you don't have such a scope.

The hum may be coming from the transformer because it employs preregulation. A sort of switchmode technique. An SCR connected to the transformer primary that regulates the output voltage of the transformer to just a few volts more than the set voltage. The series pass transistor then finally regulates the voltage down to the set voltage. That way less power is dissipated on the series pass transistor, less heat is generated and the efficiency is higher. Most of the older Sorensen and Hewlett- Packard supplies used this system. Imagine the filtering needed to get the noise to less than 100 microvolts!

What are the two recessed banana jacks beside the outputs? Are they the four wire sense connections?

If you can, I would love if you took a few pictures of the insides.

[insurgent]

Hey Tekfan,

I think the only reason this didn't disappear much faster was that they had the manufacturer name on the listing as "Soresn"! I'm definitely glad I spotted it because I was about to pull the trigger on a B&K  1650. Would have been a great beginner supply but I think this was much better. Much cooler looking if nothing else

I actually had a typo in my original message as I was thinking the noise measurements were good at < 1mVrms for a 20 year old supply and my impromptu measurements. I could be misinterpreting the reading though!
I looked around at differential probe/amps but they are out of my price range at the moment. I'll keep an eye on EBay/Craigslist for Tek 7k series though. I may be able to score a cheap isolation transformer at a local electronics surplus place.
Now that I have the Agilent U1253A Grainger special, I've verified that the voltage and current are rock solid and the displays (except for the two wandering ones) are pretty accurate.

You are right that the two recessed banana jacks are for the remote load sense connections which I haven't had time to play with yet. It also has the remote analog programming card installed which will be a great time-waster down the line.

I'll definitely post pictures of the inside for you. Hopefully this weekend or when I get time to take it apart and check out the two wandering displays. I'm definitely hoping they are an easy fix. I'll also be checking for issues that Dimlow suggested wrt to ESR and leakage.

Take care!

[tekfan]

You can construct an Isolation transformer with two IDENTICAL transformers by simply connecting both of their secondaries together. It doesn't matter what values they are as long as the primaries are wound for 120 VAC and the secondaries have the same rating.
That way you get the same voltage on the output as on the input.