That looks absolutely spot on so far Tony. ?
Hello all,
So, finally, it's finished!
Below I have some pictures of the final construction. I was able to run tests with some NOS caps I have to hand. I was glad to say that there were improvements to their ESR figures, so I'm on the right track.
So, the finished article ....
Control panel.
Front.
Side.
Rear.
Now, to my first customer 🙂
I've a 220uF @ 25v cap that's been sitting in my box for 15 years.
Here goes....
Connected in the off and discharged position.
Drawing 4mA at 25V
70uA
4uA
ESR improved from 0.212r to 0.209r. Not much, I know, but it was a good cap anyway.
Next was a 47uF @ 250V cap that was about 5 years old.
Here, the neon is flashing, whilst the cap is drawing about 1.5mA.
At 1mA, the neon is out.
Lastly, after about 15 minutes, the leakage was down to about 4uA.
The ESR improved from 0.265r to 0.261r. Again, not much, but it was already OK.
I'll be abroad for a couple of days, but when I get back, I'll draw up the circuit diagram and show the internal construction.
It's been lots of fun, designing, experimenting and building this reformer. Hopefully I can get to use it often 🙂
All the best ,
Tony
Very nice job indeed....looks just the ticket and a very useful device too. ?
Just an idea, but it might be nice to use Letraset for the legends on the fascia and a coat of lacquer to protect it. I did that with the pre-amp I made.....
Thanks for your reply Andrew,
The decals on your preamp do look nice 🙂
I might try to find a way to do something like that when I get back. The next project is looming though. There's a KB LR10 radio with my name on it, coming in the next few days.... More fun to be had 😉
All the best,
Tony
Well I'm sat in the pub so excuse my short reply.
What a cracking job, well done !
Just a quickie.
I've made up the circuit diagram for my reformer now, as below.
[edit... I saw a mistake with it, so quickly fixed that!]
I can now start restoring the LR10 that landed on my door step on Saturday.
Night night all,
Tony
Hi Tony,
Couple of query’s , it’s early so hopefully I get the right.
1. There is an adjustment for 500v at point A, you have 5x100v Zeners, were the Zeners not accurate enough?
2. Point A had me searching, it took a few seconds for my brain to realise that the big letter A in the top left hand corner was not point A. Is it worth relabelling the meter, or perhaps it’s me who is just slow these days.
I said it was early for me.
Certainley a professional job, very nice.
Hi Frank,
It turns out that the voltage creeps up after a few minutes, even with the zeners in place. As the resistor chain was heating up, the HT was going up too. In the original circuit, I had to constantly readjust the 500v trimmer, to avoid overvolts and it could take 20 minutes to settle down, even without a load. The zeners help regulate that aspect, so that once the 500v trimmer is set, then the HT remains stable. You still have to wait a few minutes and readjust, but it's better than having to readjust all the time.
I see what you mean about the big letter A... I can rename the offending test point Point F, in your honour 😉
Again, it turned out that there were very few ammeter or voltmeter symbols in the EasyEDA library. Believe it or not, those two symbols were the best of a bad bunch.
Thanks for your kind words and feedback. Most appreciated.
All the best,
Tony
There you go Frank 🙂
Fame at last ?
Might I suggest a minor modification?
You have arranged for the reforming supply to have a sensibly high resistance, which will limit the current in the event that the CUT goes s/c whilst reforming. However, if this happens, the s/c current flows through your meter which, with Sod's Law in full effect, will probably be switched to 100 microamps at the time. There is a risk that the meter will sustain damage in this event. Decent panel meters are not the cheapest of components, even if bought second-hand through ebay.
You can prevent this by connecting a suitably-chosen zener, or perhaps one or more rectifier diodes, across the meter terminals so that, in the event of the f.s.d. voltage of the meter being exceeded, the zener/diode string conducts and limits the extent to which the meter can be overloaded.
Of course, for the detail of this, you need to know the f.s.d volt-drop across the meter - which is equivalent to saying that you need to know the movement resistance, but you will need to know that anyway, to calculate the values of the shunts (your R16 - R19).
Might I suggest a minor modification?
...a suitably-chosen zener, or perhaps one or more rectifier diodes, across the meter terminals so that, in the event of the f.s.d. voltage of the meter being exceeded, the zener/diode string conducts and limits the extent to which the meter can be overloaded.
Hi occiput,
You make an interesting point, which I did think about. I think from memory, the ammeter has a resistance of 1k6, which would only need a voltage across the meter of 0.16v at 100uA for FSD. Too small for any diode to forward bias.
What I settled for is disciplined operation of the SW3. So, for instance, on voltage ranges above 100v, set SW3 to the off position until the neon stops flashing. The next position of SW3 is 100mA, which in the normal course of events is not needed.
The basic idea is to check the leakage, switch the meter off while reforming and check again later. It might be an idea to put a push-to-read switch in to isolate the ammeter, but you would still ideally have to have SW3 in the off position first when starting any readings.
One could use a second voltmeter across a shunt resistor, as in the original diagram, which could then be protected by diodes, but again, what happens at low voltages?
Slightly off piste, I can remember Samsung VCRs that used 3 series connected 1N4001 diodes to act as a very low voltage zener diode, when forward biased. This was used to regulate capstan speed. The usual mode of failure was one of the diodes not conducting when it should. The way to test for that was to read the voltage drop across each diode. If one was faulty, there would be about 12v across it and the capstan would zing along!
In case anyone's interested, the 5 Zeners have about 525v across them. Lowering the dropper resistance only increased the voltage and made the Zeners run hot. Raising the dropper resistance by too much effected how much current could be delivered, so there was a compromise to be had.
R19 is not required for my setup, because the ammeter is already rated at 100uA FSD.
Thanks for taking time to help and make suggestions to improve things. It is one of the nicer features of this place 🙂
All the best,
Tony
It was common practice with analogue multimeters to connect a pair of 1N4148 or similar diodes in parallel with the meter movement with 1 diode cathode to negative meter terminal, 1 diode cathode to positive. This ensured that a diode would conduct under overload with either polarity.
As occiput suggested, it would be wise to do this in your application.
You could even try using schottky diodes which have a lower forward voltage than normal silicon junction diodes, however, you wouldn't want the diodes to start to conduct before the meter reached full scale deflection as this would affect the accuracy of reading.
John
You make an interesting point, which I did think about. I think from memory, the ammeter has a resistance of 1k6, which would only need a voltage across the meter of 0.16v at 100uA for FSD. Too small for any diode to forward bias.
Any Si diode, perhaps.
Sounds as though an OA of some sort might be a good selection - maybe two in series.
As previously mentioned, back-to-back Schottky diodes - they typically have a 0.35V forward drop.
Thanks for the replies.
I forgot about the OA95 diodes 🙂 It was another late night though!
When I get time, I'll experiment with some OA95s to see if they'll do the trick. Right now, I only intend to use the ammeter for spot checking the leakage. At all other times, the ammeter is shorted out. In any event, the neon will be permanently on (for any voltage above 100v) if the CUT shorts out.
All the best,
Tony
Hello all,
So, a little update on the Reformer Of Boing!
Today, I hooked up with Marc, Mrs Marc and Colourstar. Whilst I was there we got the reformer out on a patient TV (a BRC 1400, I believe) to check the smoothing block.
Well, the reformer worked well and after a hesitation, so did the block. One of sections got down 60uA at 275V (250V setting trimmed up to 275V, using the Trim Voltage pot). After about half an hour, all sections proved to be fine 🙂
Things run a little warm inside the enclosure, so I'm in the process of fitting a little cooling fan to improve airflow.
A big 2 thumbs fresh to Marc for providing me with a couple of OA81 diodes to protect the meter ? ?
I've attached some photos... Sorry about the camera shake... I didn't realise until I got back home. Anyway, here they are 🙂
All the best,
Tony
I have to say that it's a cracking re-former that Tony has put together, well built and plenty of range. I might have to consider a new build myself....once I get my workshop functioning again !
Great bit of kit. ?
Can I ask is the circuit your own creation or an adaption of one?
Hi Chris,
The circuit is an adaption. I can post up the pdf later if you want.
I would appreciate seeing the .pdf as well whenever convenient. I bought a load of NOS electrolytics and I certainly want to reform/test them before use. I need to build a reformer...
Thanks
Peter
BTW a very workmanlike piece of kit that’s earning its keep and a great write-up. I already feel a hand reaching for the iron to build my own...
Sorry had not spotted the circuit pictures when I asked for the .pdf.
Peter
