Resurrecting HP 5245L frequency counter E515 prefix

[factory]

The remains of this HP 5245L were given to me with some TE I bought, it's one that was made at the HP Bedford site, prior to the Scotland move, there seems to be very little remaining from the first few years of HP UK.

As I got it, it was missing all the removable PCBs & the case covers, I thought of using it for parts, as the cable loom could have been used to update my 5243L for the later plug-ins. The side frame got used on something else, I can't remember what, the crystal oven control board got disconnected when I was having problems with the 5243L and it had been sitting in this state for a few years, I just couldn't bring myself to scrap it.



A few weeks back I found someone offering a pile of boards/parts from these counters, I bought most of the missing boards needed (except the decade divider boards), I also had a couple of display boards from Sphere in Canada too (minus the numerical readout devices), finally I could attempt to resurrect this counter.



David

[factory]

First job was to give it a good cleaning, then I reconnected the oven controller board & added an extra wire for adding a fuse to the heater output, to prevent transformer failure if the crystal oven heater winding shorts.
At the same time I installed two sockets, for the optional extra decade divider boards (100 & 1000 second gate time).



Next was to install all of the boards, I checked the rectifier board & found several bad diodes, good job I did this before installing it, I've borrowed one from another counter for testing. It was then brought up slowly using various lamps in the lamp limiter, the power supplies were checked, two were OK, but the +20V had a large ripple on it, I wasn't surprized to find the 1500uF 15V capacitor C10 was open circuit (the exact same one, that was open in two different 412A voltmeters), the others all tested fine. A temporary replacement 1500uF 25V has been fitted to the card edge connectors underneath, will re-stuff the original in the future.



David

[factory]

As can be seen in the previous post, only one of the seven numerical readout bulbs I got in the parts lot is working, hardly surprizing given they were from a coastal area, these end view types seem to have quite a high failure rate IMO, all with cracked glass from pin corrosion (oxide jacking).

I'm ignoring this at the moment, as there were other faults to fix first, firstly the gate light is on all the time, secondly the timebase output on the front was only giving outputs at 10Mc & 1Mc (this older counter uses the cycle 'c' symbol not 'Hz').

The timebase problem was looked at first, after a few false leads due to using a paper manual for the 1964 version, apparently the A26 5243A-65F oscillator board is quite different to the A26 05245-6013 in the 1965 version, they are completely incompatible due to wiring differences. This counter uses the newer of the two.

Next was to investigate the decade divider chain, this takes one of the three 1Mc outputs from the A26 oscillator board & divides it down via a higher speed board A28, then six to eight lower speed boards (A29-A34 & A50-A51), each board divides by ten.

The 1Mc was getting to the A28 board, but no output, checking each pair of flip-flop transistors it was only getting past the first pair, borrowing another board confirmed this fault.



The leads of the two transistors (Q3 & Q4) were quite rusty, they were removed, a lead fell off one, the other had half of the junction open internally, two replacement 1583-0034 were taken from one of the Sphere spare boards, this fixed the board and I finally had output on all the timebase settings.



David

[factory]

And the gate light fault, this was lit continuously, turning the sample rate/ hold control made no difference, the manual states to check the A22 gate control board.
The signal was getting past Q1, but everything else seemed to give no signal, just DC levels, the frequency, start & stop on the function switch were giving strange levels, it seemed to switch high & low turning between the three functions, but stuck in one state with just start & stop.

Something was very wrong with this board, I was measuring a lot of short circuits on the 2N708/2N709 transistors with a DMM on diode check, pulling them one by one, until only one of the seven small NPN silicon transistors remained, the removed ones all checked either shorted or as Zener junctions on the tester.



Searching through my spares I managed to find enough good used ones, to replace the six failed transistors, after this I now had a fully working gate board and usable one digit (ATM) HP 5245L electronic counter.



Next will be to try & find some more numerical readout bulbs and replace a few failed symbol & decimal point indicator lamps.

David

[jez_145]

Hi David,

You're a brave man!

I don't know if you are aware of this but there's a chap called Marc Verdiell that runs YouTube channel curiousmarc (and also his related web site). He has a great interest in all (vintage) things Hewlett Packard. Recently he did a lot of work restoring a 5425L counter, and its associated plug in modules, which may be of interest to you. Either way, he has an excellent channel and is well worth a look (one of my favourites!)

curiousmarc YouTube channel is here:-

https://www.youtube.com/c/CuriousMarc/videos


Playlist specific to the HP 5425L restoration is here:-

https://www.youtube.com/watch?v=B-4c...gdKCaUPY1zaPh_


Jerry

[factory]

Thanks Jerry, either I'm brave or it's insanity but these early UK made ones aren't very easy to find these days, too many have been parted out by people not interested in our historic test gear. Oddly there seems to be more UK made HP TE appearing for sale in Germany than here.

I've watched the videos on the 5245L, it's plug-ins & the 5327B that Curiousmarc has repaired, they are excellent videos.

David

[cmjones01]

This thread has reminded me of my own 5245L which is in storage because there's sadly no room for it in my lab. I got it from the Pye TVT factory in Cambridge in 1987 and it's great, though the crystal oven has been troublesome. It's burned out at least two oven mains transformers in my ownership, due to damaged wiring in the oven latterly caused by the Germanium power transistor that feeds the oven going leaky and thus cooking everything. I had to hack the insulation out to repair it. I replaced it with fibreglass loft insulation so the heater has to work harder than it used to!

I added a confidence check feature: switching to the sensitivity range formerly known as 10V now displays the oven temperature in tenths of a degree C.

I don't have a good photo of the back panel so can't remember the full serial number right now, but it's UK made and ends 02543.

A few years ago I scratched the itch to acquire a few plugins useful to me: the time interval counter and video amplifier (Nuvistor!) are neat, the voltmeter is surprisingly good with a full 6 digits of resolution, and the 5257A transfer oscillator is just so cool in spite of being fairly hard to use.

Chris

[trobbins]

I repaired a Systron Donner 6250A of what looks like similar vintage and with 1Hz resolution for 10MHz, but a much better storage history than those photos show since the 1970's. That was the first step that then led to how to calibrate such a counter. Luckily I came across a cheap diy GPSDO that easily provided 0.1Hz resolution, so I was in calibration heaven (not a typical outcome for some metrology pursuits).

[6AL5W-Martin]

good luck with the counter.
And let us know what°s going on in the restoration.
You are not allone

greetings
Martin

[factory]

Originally Posted by cmjones01

This thread has reminded me of my own 5245L which is in storage because there's sadly no room for it in my lab. I got it from the Pye TVT factory in Cambridge in 1987 and it's great, though the crystal oven has been troublesome. It's burned out at least two oven mains transformers in my ownership, due to damaged wiring in the oven latterly caused by the Germanium power transistor that feeds the oven going leaky and thus cooking everything. I had to hack the insulation out to repair it. I replaced it with fibreglass loft insulation so the heater has to work harder than it used to!

I added a confidence check feature: switching to the sensitivity range formerly known as 10V now displays the oven temperature in tenths of a degree C.

I don't have a good photo of the back panel so can't remember the full serial number right now, but it's UK made and ends 02543.

A few years ago I scratched the itch to acquire a few plugins useful to me: the time interval counter and video amplifier (Nuvistor!) are neat, the voltmeter is surprisingly good with a full 6 digits of resolution, and the 5257A transfer oscillator is just so cool in spite of being fairly hard to use.

Chris

That one is a little newer, it has the HP logo badge, 'Hz' symbols instead of cycles and the vernier adjustment for sensitivity. I had a look through my saved pictures and found a U717 prefix (sn 02477) that is close to your serial number, made at the South Queensferry site.
Also found an early UK one E328 sn 00112, sold in Germany, hope it's survived.

I need to look up your repair thread, want to find out the transformer rating and figure out the most appropriate fuse value, to add in series with the heater.

You have a few of the useful most plug-ins, there is also a later FET version of the 5261A video amplifier plug-in, I have one here with a few of others shown below, also pictured is my later 5245L prefix 1840A, it uses the newer 10544 oven.
Still haven't found a 5253A, 5254C, 5263A & 5268A. There is one that got recalled, the 5259A frequency converter, no idea if any survived, also don't know if 5266A got allocated to a plug-in.



Something a bit rarer, a H16 5243L counter, it's 20Mc maximum with the 100 second gate time option, had to import this one from the US, haven't seen one sold here for many years. It's pictured with a 512B frequency converter, the Systron DMM was monitoring the HT rail during capacitor reforming.

Originally Posted by trobbins

I repaired a Systron Donner 6250A of what looks like similar vintage and with 1Hz resolution for 10MHz, but a much better storage history than those photos show since the 1970's. That was the first step that then led to how to calibrate such a counter. Luckily I came across a cheap diy GPSDO that easily provided 0.1Hz resolution, so I was in calibration heaven (not a typical outcome for some metrology pursuits).

Haven't got a timer/counter from Systron, but do have a couple of frequency counters of similar age from them, plus the DMM above and an LC meter.

David

[factory]

Originally Posted by 6AL5W-Martin

good luck with the counter.
And let us know what°s going on in the restoration.
You are not allone

greetings
Martin

An update on the older 5245L repair, I've found some new readout bulbs, one is of course defective from storage for several decades, got to locate the few NOS HP ones I know I have somewhere.



Decimal point & symbol lamps are a little tricky to access, replacement is currently on hold while I figure out how to make some ordinary lamps into frosted versions, only the decimal points were frosted originally, probably will replace all of them, as the leads of three are broken due to rust, the others would probably break moving the assembly back into place.
Frosted glass paint is available, but it needs baking and I would need to buy one of those cheap toaster ovens & modify it, as people do to make cheap reflow ovens.



David

[6AL5W-Martin]

about plugins:
the video amplifier is a must have, with a simply scope probe the counter is always at work.

on my picture you see the Digital Voltmeter plugin. The accuracy is amazing. But it use only 6 of the digits bec. it`s not a 3458A