DC Power Supply



#2

I have always gotten good and interesting advice from the folks who post to this forum, and this is at least peripherally related to HP calculators, so...

I am in the market for a variable voltage DC power supply for bench testing calculators and other small electronic devices. I want to be able to plug it into a wall outlet (in the U.S., so 110-120 V AC, 60Hz). I also do not want to pay an enormous sum for it ($100 or less would be preferrable).

I have found a company called Mastech which sells a unit with adjustable output (0-30v,0-3amp) and digital readouts for around $100.

Does anyone have any experience, comments, etc. regarding this company's power supplies. Any other suggestions for a power supply?

Thanks.

Take care.

Wayne


#3

Hi, Wayne;

I built many worbench power supplies for my own use and some for commercial applications (private users) and also for a particular employer.

What bothers me most about variable DC voltage sources is the possibility of fault. In most cases, protaction against faulty operation is a lot efficient, but when we are dealing with a 0 to 30V output, I am not sure if the output protection will be so accurate to go further.

The fact is that when I'm dealing with too low power consumption devices (calculators, in this case), I decided using rechargeable battery packs and keep a multipurpose charger. Batteries are my "power supply", and the worst case is that they may be low in voltage, never up. No transients, no AC-DC fault, no bad contact when adjusting voltage (variable resistors, unless it is digital). And the calculators are actually designed to use batteries as power supplies, so it's not such a waste. I build a pack with four regular D-size batteries (not rechargeable) to use with my HP41 (fullnuts with external DC access) when reading cards or using wand or IR module (printer). It's about a year since I built it, I read a lot of cards, tested and loaded many of my own programs in barcode printout (using 41uc.exe)... and the batteries show no fatigue. Amazing. As I'm dealing with LCD models, they already use regular 1.5V batteries, and their average current is low enough.

If I'm dealing with older LED models, I use 1.1Ah NiCAD or 1.6Ah NiMH AA-size rechargeable (I bought four Panasonic 1.95Ah and I intend using three of them to power an HP75D). Anyway, about two days ago I found and bought four NiCAD D-size marked 5Ah! That's a lot more than enough for any portable application, devices included, I guess.

This is what I can tell you based on my own experiences. I never used a DC power supply with any calcualtor I ever repaired. I guess batteries are safer...

Cheers.

Luiz (Brazil)


#4

Thank you, Luiz.

I had not really considered the effects of minor variations in the low voltages needed for calculator testing. And once again I appear to be over-complicating things.

Perhaps I will go ahead build my own "bench" supply using batteries. D cells are cheap. And I've been using 2 slightly used AA cells with quite a bit of success in my old 30 series LED calcs. I can probably throw together some kind of holder with leads on it to contain those as well. It'll be considerably less money than buying a commercial grade power supply.

Take care.

Wayne.

#5

I use a Velleman kit bench power supply. It was _not_ cheap (about \pounds 220 I think!) but it's rated from 0-30V at up to 8A _continuous_. Way overkill for running handhelds, but I do a lot of other electronics besides, and I have needed a fairly high current supply for that.

One thing that IMHO is essential is an adjustable current limiter. You don't want the current supplied to the calculator to go through the roof if there's a fault (Topcats, in particular, will draw a couple of amps in an attempt to pull the supply voltage down if the power converter doesn't run). A NiCd pack will easily supply this sort of current, and it may cause damage elsewhere.

FWIW, I've seen TI calculator (SR51A IIRC) with melted tracks on the PCB due to the current supplied by its NiCd pack when something shorted out!

I have no experience of the supply you mention, but I have had experience of some of the cheaper supplies available in the UK. They ranged from positively lethal (no earth (ground), no short-circuit protection, etc) to just useless (one I remember would put a spike of about 35v on the output during power up / power down and would routinely cook whatever it was connected to as a result).

#6

I don't know anything about that power supply specifically,
but if the current limit is adjustable, it should be OK.
The current limit serves to prevent damage to the DUT (device under test) if there are fault conditions. For instance, if you know that the DUT should draw no more than 200 mA at 2.5V and set the power supply to those limits, but there is a bad component causing it to draw more, the power supply will go into current limiting. This will prevent other components of the DUT from being stressed beyond their ratings.

A current-limited DC power supply has two independent limits, voltage and current, and two mode. Normally the power supply operates in constant voltage mode, where it provides the set voltage at a very low impedance, so it maintains voltage regulation with a variable load. But if the load exceeds the current limit, the power supply switches to current limit mode, in which it has a higher output impedance, and regulates the current. This results in the voltage being lower than the voltage limit. Many bench power supplies have "CC" and "CV" indicators, though the same information can be inferred from the meters.

If you were to draw a voltage-vs-current graph, it would have line segments at V=Vlimit for C=0 to Climit (CV mode), and at C=Climit for V=0 to Vlimit (CC mode).

Of course, this description is for an ideal power supply. A real power supply will deviate slightly due to imperfect line and load regulation and finite transient frequency response. In particular, highly inductive or highly capacitive loads may significantly reduce the regulation performance of the power supply. (For testing calculators this should not be an issue.)

For non-printing LED models, set the current limit to
about 200 mA. Printing models need much more current while they are printing (>500 mA), but they should initially be tested without printing at the 200 mA limit. Once you've determined that the non-printing functionality works correctly, increase the current limit to 600 mA and test printing. Models equipped with a card reader will also draw more than 200 mA when the card reader is in operation; I haven't yet done any characterization of the current requirements.

The later all-CMOS calculators (41C family, Voyager family) should be tested at a much lower current limit, except when using the 82104A card reader. However, I normally just test those with disposable alkaline batteries.

If the power supply doesn't have a direct method of setting the current limit, you can do it by shorting the output (which puts the supply into constant current mode), then adjusting the current limit knob until the measured current reads the desired limit.

I use an HP E3631A programmable triple power supply, but that is much fancier (and more expensive) than what you need.


#7

I forgot to add that you're actually better off with a power supply with ANALOG meters rather than digital. Though they might not have as much effective resolution, they are far more useful in the sense that you can see exactly what is going on with a quick glance.

The reason I have the E3631A power supply I previously mentioned, which has digital meters, is that I needed a programmable power supply, and there aren't (AFAIK) any inexpensive ones with analog meters. I routinely use the E3631A in combination with a separate analog meter.

Edited: 27 Apr 2004, 8:32 p.m.

#8

Wayne,
Have you considered getting a vintage HP DC bench power supply? There are very nice, analog meter units at ebay, many in what appears to be good shape. I am waiting for one to be delivered to my doorway here in Brasil, but Santa thinks I´m not a good boy. In my opinion, a vintage HP power supply would be THE choice for testing vintage HP calcs.

Renato


#9

Hi Renato, all;

I did not consider this option (neither was known about this fact), but after Eric's wise, very good remarks about current limits, I'd vote for these vintage HP, too. I never found any calculator in the condition Eric mentioned (severe short circuit), but it is indeed something to be taken too seriously. (thanks, Eric)

Ahn... Renato, if Santa knocks at your door, you know my actaul address.. don't you? ;^)

Cheers.

Luiz (Brazil)

Edited: 28 Apr 2004, 12:35 a.m.


#10

Hi Luiz,

I´ll include your request in my next letter to Santa.
Will you come around here ? J. Ernesto and me would be very happy to host you.

I had a interesting situation with a HP-27 once - it was Ok when it arrived, but quitted working in a few days. I opened it and discovered that the copper PCI trace to one of the battery terminals had become a fuse - it melted and opened circuit. It was a easy fix, but just imagine what would have happened if I tried to charge batteries..


Renato


#11

Wow!

I cannot imagine. I don't like horror movies, mostly when the good guys die at the beginning...

I'll let you know as soon as I'm planning to go. it seems a new condition is about to take place, so...

Cheers and best regards.

Luiz

#12

I use a Harrison/HP 6101A thai I paid 5 bucks for...


#13

David,

5 bucks ? If there are more available, I could save my xmas wish (and Luiz too). Would there be chance for two units be sent to Brazil ?

Renato

Edited: 28 Apr 2004, 9:01 p.m.


#14

I bought the last one. Until last week he had a 6102A. Shipping to Brazil would probably be a whole lot. They are kinda heavy. They do show up on Ebay quite a bit. I would like to get the manual for it someday. The voltage meter needs to be calibrated.


#15

Here's a great, small and very lightweight power supply that's perfect for working with all the HP calculators. I've used these supplies many years ago in a research lab and they're exceptionally reliable. There are several other supplies in this series that go to higher voltages but won't supply enough current to run the printing calculators -- this one just barely makes it on the HP-97.

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=58286&item=3812004606&rd=1&ssPageName=WDVW

$30 is about the going rate on these.


Possibly Related Threads...
Thread Author Replies Views Last Post
  HP 9845 option 200 power supply Dave F 2 615 05-30-2013, 08:10 AM
Last Post: Dave F
  Precision DC-50 Mic 3 614 05-26-2013, 01:27 PM
Last Post: DavidM
  HP-87 power supply repair philippe_elie 0 333 11-18-2012, 11:12 AM
Last Post: philippe_elie
  (OT) Discovery over DC David Tellet 9 970 04-20-2012, 01:24 PM
Last Post: Richard Ottosen (Denver)
  strange classic power supply (pics) Alberto Fenini 4 756 01-16-2012, 02:06 PM
Last Post: Ethan Conner
  Bench Power Supply aj04062 1 341 11-12-2011, 01:59 PM
Last Post: Eric Smith
  Canon Pocketronic - Power Supply Joerg Woerner 3 497 10-26-2011, 09:48 AM
Last Post: Katie Wasserman
  3V Power Supply Lode 3 509 08-31-2011, 02:41 AM
Last Post: Marcus von Cube, Germany
  HP71b - which power supply Alberto Fenini 6 617 08-18-2010, 12:32 PM
Last Post: Alberto Fenini
  COMPU-LS and COMPU-DC Mike 6 693 07-17-2010, 07:34 PM
Last Post: Forrest Switzer

Forum Jump: