BlackStar 4503

[synx508]

Here's another DMM that I'd been curious about for quite some time. It has plenty of features but fairly basic performance.



This had two major faults, the NiCd battery had failed and leaked and the power switch is very reluctant to switch on without a lot of alarming arcing.

I quickly removed VARTA, destroyer of worlds, replaced with two NiMH AA cells in a holder, though I might change this to AAA when I get a replacement battery holder. The PCB needed some cleaning but the leakage wasn't as bad as I've seen in other equipment. I changed C21, a 560pF capacitor and adjusted the parallel trimmer to accommodate the slightly different value of my replacement part.

I've left the switch for now but I need to remember not to turn it off, too.

I fed it a pile of resistors, then some outputs from my bench supply and HP 3325A+HV option, measured using the HP 3456A for reference. I didn't succeed in the current calibration, it kept saying BAD, but it wasn't far off, everything else worked.

It's not a bad meter but the AC ranges take several seconds to settle and that's a bit annoying. I think it's doing software averaging. I really like the big LED display and assuming it's reliable it might even become my daily driver on the bench (the Fluke 45 is becoming cantankerous and its VFD is fading).

Another bad point is disassembly, which was horrible due to having to remove hard-to reach plastic and metal nuts, but I'll forgive it that if it continues to work for a few weeks.

[synx508]

DC Current calibration is now complete, it really, really needed the screening metal to be in place.

[AVGresponding]

They're a decent meter. The SRAM battery is very problematic though; I'm contemplating binning off rechargeables on mine (I did do a 2x AAA mod but prolonged lack of use results in cal loss) in favour of hot-swappable lithiums of some sort.


EDIT: Oh, btw, if yours has the GPIB, TheHWcave is looking for a ROM dump (I don't have the means).

[EC8010]

Excellent; another aged meter brought back to life. And it looks like its display goes to 3.9999, making it quite reasonable accuracy. Useful.

[MED6753]

You guys are making me look bad. I gotta find some decent TE to work on. Been very lean here lately.

[synx508]

Excellent; another aged meter brought back to life. And it looks like its display goes to 3.9999, making it quite reasonable accuracy. Useful.

Yes, it's a 40,999 count design which makes it fine as a basic bench meter. Apparently there's an earlier 4503 that is 20,999, which must be a bit confusing.

I've decided against replacing my Fluke 45 with it, because the BlackStar doesn't have a frequency counter and I use that feature quite regularly (fed with the sync output on my 3312A).

It has 10MΩ input impedance on voltage ranges, though, which is a bit disappointing. The DC accuracy is a fairly poor ±0.03% reading ±0.01% full scale.

[synx508]

They're a decent meter. The SRAM battery is very problematic though; I'm contemplating binning off rechargeables on mine (I did do a 2x AAA mod but prolonged lack of use results in cal loss) in favour of hot-swappable lithiums of some sort.


EDIT: Oh, btw, if yours has the GPIB, TheHWcave is looking for a ROM dump (I don't have the means).

I do have GPIB on mine, it seems to work and I've figured out most of the commands though I'm waiting for KO4BB to start working again so I can download the full user manual - I have a service manual but it doesn't have the full list of commands in it. KO4BB also has firmware and I think it's a newer version than I have, but again, not currently accessible. When I change the power switch, maybe this weekend, I'll read the ROM and put it on my website. It has a Hitachi 6303x CPU, which is an enhanced clone of a 6800 with some peripherals, seems quite good and I'm tempted to try to add features…

[synx508]

Here's a link to a permatemporary web page with a link to my extracted firmware. https://mossyvale.co.uk/te/

[synx508]

KO4BB was working recently and I've downloaded the firmware and operating manual now.

[Specmaster]

The meter is a really useful one and is one of my favourites. When I got mine, it was way out of spec on all ranges as the VARTA had leaked and the resultant flood of electrolyte had reached the extremely high impedance front end and it took many attempts at cleaning the board, but I'm glad I did.

The calibration was a bit of a pig to do, but with perseverance and adaptation, I managed to get it reading close to my +/- 1 or 2 mv (at the time) new Bryman BM867s and my Flukes, which all agreed with my calibrators.

[AVGresponding]

With battery backup on my mind recently (planning a dual battery mod for my 3478A's, to enable safe hot-swapping), I decided to have a look at the 4503 schematic, to see if there's any possibility for improvements.

There's a few oddities, or bad design choices, depending on your POV. The battery is directly connected to the non-inverting inputs of the TL072 op-amps used in the RS232. Ok, these are high impedance inputs, but that's still a poor choice when you have a regulated 5V supply available.

Secondly, the VBACK rail, which is DT OR'ed with the 5VDig_reg and VBATT rails acts as an external pull-up for two of the SRAM active-low pins, OE and CS1, as well as part of the CPU reset circuit. There's a 22kΩ resistor in series, effectively limiting maximum current to 100µA, and as they are active low there'll be internal pull-ups, probably. It shouldn't* be a problem, it's just not how I'd do it.

Lastly, the charge circuit is terrible. It's overly simple, and while you can probably get away with it with NiCd, it'll kill a NiMH quite quickly imo, as the OCV is going to be at least 3.6V, far too high for a 2S NiMH.

So, my plan, is to carry out the same type of mod as on the 3478A. LiSOCl2 and LiMnO2 are both suitable, as the SRAMs in question will work down to 2V, giving you plenty of warning once the cell drops below 3V. I'm using 1/2 AA battery holders, giving a capacity of ~900-1200mAh, which should be plenty.

Most of the figuring out on this is going to be where to put the battery holders; probably not a bad idea to provide an external connector for monitoring the VBATT as well.