It is currently Fri Jun 22, 2018 9:40 pm

Testing 5U4 rectifier

 
Posts: 706
Joined: Sat Jan 07, 2012 10:21 am
Location: Sleaford lincs

Testing 5U4 rectifier

Post by Dr Wobble » Sat Mar 22, 2014 12:01 pm

I want to test a 5U4 rectifier but am a bit unsure how to wire it up,as it has no cathode. I know the filament acts as the cathode and the HT positive would usually be taken off the filament center tap,with the 0v off the center tap off the HT secondarys .

On my homemade VT I would normally wire up a valve with 0v to the cathode. If I wire the rectifier up with HT + to pin 4 anode, feed the filaments with 5v and wire 0v ( usually to the cathode) to pin 8 would this work? Or would it be better to re-wire specificly for a rectifier,not sure if my transformers have a center tap. I dont want to blow anything up so any advice would be most welcome.

Andy.

 
Posts: 86
Joined: Sun May 13, 2012 3:15 pm

Re: Testing 5U4 rectifier

Post by boyblue » Sat Mar 22, 2014 12:31 pm

Hi Andy, I,ve never seen a 5V rec heater winding with a CT, you usually take the HT from either heater pin. I don't think it matters which htr pin you use. I,ve always thought that this setup would introduce some hum but its always been ok for me.

Peter

 
Posts: 2513
Joined: Sun Jun 10, 2012 7:23 pm
Location: Basildon

Re: Testing 5U4 rectifier

Post by crackle » Sat Mar 22, 2014 12:39 pm

Hi Andy
I think you will need to limit the current flowing between the anode and cathode. Not sure what value but if it was rectifying 250v then a 1.2k resistor should limit the current to 200mA

 
Posts: 917
Joined: Sat May 26, 2012 5:41 pm

Re: Testing 5U4 rectifier

Post by Valvebloke » Sat Mar 22, 2014 1:30 pm

boyblue wrote:Hi Andy, I,ve never seen a 5V rec heater winding with a CT, you usually take the HT from either heater pin. I don't think it matters which htr pin you use. I,ve always thought that this setup would introduce some hum but its always been ok for me.

Peter


Here's one https://www.edcorusa.com/xpwr215.

As you say, if you take the HT from one end or the other of the filament then the voltage across one of the two diodes (measured from the anode to the centre of the filament) will always be 7Vpk higher than the voltage across the other one. This will put a 50Hz ripple onto the HT as well as the 100Hz one that's always there and will disproportionately stress the diode which is seeing the higher voltage (and therefore passing more of the current).

One way round this is to put a voltage divider, made from two equal resistors in series, across the 5V winding and take the HT output from the centre point. Of course this combination of resistors will be in series with the rectifier (in this case the effective series resistance will the value of the two resistors in parallel, since each will be taking roughly half the HT current). But we quite often need to add a resistor in series with the rectifier anyway. So we can kill two birds with one stone this way.

VB

 
Posts: 6972
Joined: Sun Aug 14, 2011 11:02 am
Location: Co. Limerick

Re: Testing 5U4 rectifier

Post by Michael Watterson » Sat Mar 22, 2014 1:33 pm

put 5V on the filament.

Then a DMM will read a few K one way and open the other.

Connect a current limited bench PSU to anode and filament. + to Anode. See what Anode voltage gives the max rated load current. If you can't reach the rated current even at 30V, the emission is poor.
What ever anode voltage gives rated current is telling you approximate effective resistance Va / Ia

No HT or AC needed to test a Signal or Power Valve rectifier. Only heater supply and Low voltage variable PSU.

Do NOT use HT supplies, AC Transformers or Load resistors to test a rectifier emission.


To test leakage / breakdown:
With heater on, connect + of leakage tester to filament (or cathode on indirect type) and -ve to anode. There should be no current. Most mains Rectifier valves should pass even on a 600V capacitor leakage tester (which will have 1M to 3M internal series resistor to limit current).

 
Posts: 917
Joined: Sat May 26, 2012 5:41 pm

Re: Testing 5U4 rectifier

Post by Valvebloke » Sat Mar 22, 2014 2:55 pm

Michael Watterson wrote:... No HT or AC needed to test a Signal or Power Valve rectifier. Only heater supply and Low voltage variable PSU.

Do NOT use HT supplies, AC Transformers or Load resistors to test a rectifier emission.
...


The 5U4G is rated at 225mA average current (presumably 115mA or so per diode) but 675mA peak. Each diode might have enough emission to deliver 115mA, in which case a DC-only test would say it's OK. But it might run out of steam well before it gets to 675mA, in which case it might not work properly in the circuit for which it's intended. Or is it OK to test each diode all the way up to 675mA DC ? (Not for long, I fear ...)

There's a general principle that the best way to test a valve to see if it really will work properly in the circuit for which it's intended is to put it into the circuit for which it's intended. The 5U4G is intended to be an AC rectifier. I'm afraid I wouldn't trust any low-voltage DC test quite as much as I would a high-voltage AC one.

Ordinary 'cooking' 5U4Gs have recently been selling on eBay for £10-£30 each. But if you've got a nice old wide-bodied one with U52 written on it somewhere then you can add another digit in front of that. Given those sorts of values I'd want to be pretty sure of the testing.

VB

 
Posts: 4092
Joined: Wed May 02, 2012 3:35 pm
Location: Worksop

Re: Testing 5U4 rectifier

Post by Refugee » Sat Mar 22, 2014 3:16 pm

In most directly heated double diode rectifiers there is two heaters connected internally in series.
You can reverse the heater winding to add or deduct the heater voltage from one side or the other.
The amount of voltage involved is small compared with the HT winding.
If you use 10% resistors in the anodes the voltage error will be smaller than the tolerance of the resistors anyway.
I suppose that the heater could be phased to get the extra bit to add to the physical outer HT winding as the turns have further to go around the coil former.
It is nit picking really but there does indeed exist a small difference and if you want to do a perfect job it is worth considering it as an exercise in order to prove that it is not worth bothering about.

 
Posts: 6972
Joined: Sun Aug 14, 2011 11:02 am
Location: Co. Limerick

Re: Testing 5U4 rectifier

Post by Michael Watterson » Sat Mar 22, 2014 3:39 pm

Valvebloke wrote:
There's a general principle that the best way to test a valve to see if it really will work properly in the circuit for which it's intended is to put it into the circuit for which it's intended. The 5U4G is intended to be an AC rectifier. I'm afraid I wouldn't trust any low-voltage DC test quite as much as I would a high-voltage AC one.


True maybe for Mixer /Osc (especially) or RF/IF amps or high Power amplifiers. But a rectifier can be entirely statically tested. High voltage 2M Ohm limited DC to test Peak Inverse voltage and low voltage current limited PSU for Emission.

I don't know of any ordinary valve tester that measures conversion gain for Mixer / Osc. Or indeed RF gain of oscillator part. Or L.O. leakage out of RF in pin.

Yes you can wind up to peak current long enough to see if it does it without damage.

The 3 things on a rectifier:
1) PIV (leakage / flashover) (Capacitor / Leakage tester etc)
2) Forward V/I curve . Ordinary PSU.
3) Switching time / Capacitance. Only relevant only RF applications usually.

Plotting I/V roughly will tell you actually what current the emission will limit at if it's worn you get S shape curve instead of simple square law.

The Peak current depends mostly on immediate load capacitance and transformer resistance vs average load current. It's not good for any valve rectifiers to have too high a peak current, damages cathode coating. Metal Rectifiers and Semiconductor rectifiers the peak current isn't such an issue as average dissipation compared to valve rectifiers.

There is nothing magical about valve rectifiers. But most valve testers have poor instructions for testing them and essentially tell you nothing more (and maybe less) than a leakage tester, Low voltage adjustable PSU and heater/Filament supply.

 
Posts: 86
Joined: Sun May 13, 2012 3:15 pm

Re: Testing 5U4 rectifier

Post by boyblue » Sat Mar 22, 2014 3:42 pm

Hi All, it would be nice to try taking the HT from each end of the heater in turn to find out if the ripple varies, on load of course. Any takers.?
Peter

 
Posts: 6972
Joined: Sun Aug 14, 2011 11:02 am
Location: Co. Limerick

Re: Testing 5U4 rectifier

Post by Michael Watterson » Sat Mar 22, 2014 4:22 pm

I doubt it makes much difference if any in a real circuit that has two capacitors separated by choke, resistor or field coil. The ripple from the HT AC rectification is 40x more.

 
Posts: 917
Joined: Sat May 26, 2012 5:41 pm

Re: Testing 5U4 rectifier

Post by Valvebloke » Sat Mar 22, 2014 4:31 pm

Michael Watterson wrote:[... The 3 things on a rectifier:
1) PIV (leakage / flashover) (Capacitor / Leakage tester etc)
2) Forward V/I curve . Ordinary PSU ...


Could either of these things depend on the valve's temperature (=gas level) ? If so then they really should be tested with the valve properly hot, as it would be in service. I realise that the dissipation heating isn't Ohmic but I reckon that short high-current pulses might still raise the temperature more than the same average DC current.

But most valve testers have poor instructions for testing them and essentially tell you nothing more (and maybe less) than a leakage tester, Low voltage adjustable PSU and heater/Filament supply.


I fear that most valve testers are not just lacking in their instructions but are actually lacking in their ability to test large rectifiers at full load at all. Fortunately for Andy he's building his own tester, with chunky supplies, so he can do the job properly.

boyblue wrote:Hi All, it would be nice to try taking the HT from each end of the heater in turn to find out if the ripple varies, on load of course. Any takers.?
Peter


I'd be surprised if it varied much between one end and the other, although I suppose there will inevitably be some mismatch between the two diodes and this might be balanced by the voltage mismatch one way round and exaggerated the other. The basic point point though is that there's a 50Hz component there if you take the output from either end. But there isn't, with a perfect rectifier at least, if you take it from the middle.

VB

 
Posts: 6972
Joined: Sun Aug 14, 2011 11:02 am
Location: Co. Limerick

Re: Testing 5U4 rectifier

Post by Michael Watterson » Sat Mar 22, 2014 5:39 pm

But there is 200V of 50Hz or 200V of 100Hz for 1/2 or full wave. Or whatever the AC HT is. The 5V 50Hz isn't significant. Imagine no capacitor and just a load resistor.

With Low Voltage bench PSU you can make a valve rectifier hotter than most valve testers. P = V x I
It makes no difference if DC or 100Hz short pulses charging a cap. Average Power = RMS volts x RMS current.

Also for more powerful (= lower voltage drop or resistance) most valve testers don't have a low enough Va setting. 12.6 is usually minimum. Testing signal diode valves too you are interested at low currents and low Anode voltages.

Valve testers do Triodes and Pentodes (actually some only test Pentodes Triodised!). They are often poor for Hexodes, worse to useless for Heptodes /Octodes/Nonodes and very approximate for signal or Power rectifiers where a Bench PSU, V / I graph and reverse leakage test from cap tester or similar is more useful.

Triodes, Pentodes and such you may be interested in 30V to 300V range of Anode voltages vs grid bias. By definition the anode / Cathode voltage in forward "on" mode is low on a diode/Rectifier.

 
Posts: 917
Joined: Sat May 26, 2012 5:41 pm

Re: Testing 5U4 rectifier

Post by Valvebloke » Sat Mar 22, 2014 6:22 pm

Michael Watterson wrote:With Low Voltage bench PSU you can make a valve rectifier hotter than most valve testers.


I agree. Most valve testers don't test power valves of any sort, including power rectifiers, well. But a homebrew high power tester, as Andy's could be, might well be able to push them close to their actual 'in service' operating conditions.

P = V x I It makes no difference if DC or 100Hz short pulses charging a cap. Average Power = RMS volts x RMS current.


The RMS value of 100mA average current supplied as DC is 100mA. The RMS value of 100mA average current supplied as a 20% duty cycle pulse train (so 500mA peak) is 224mA. If the loss mechanism was Ohmic (I know it isn't in a thermionic diode, but indulge me for a moment) then the RMS voltage across the diode would also be raised by the same factor. So the average power being dissipated in the diode would be 2.24 x 2.24 = 5 times higher. I would call a factor of 5 'a difference'.

VB

 
Posts: 262
Joined: Wed Dec 14, 2011 1:36 pm
Location: Lynton, N. Devon

Re: Testing 5U4 rectifier

Post by Kalee20 » Sat Mar 22, 2014 7:29 pm

Valvebloke wrote:
P = V x I It makes no difference if DC or 100Hz short pulses charging a cap. Average Power = RMS volts x RMS current.


The RMS value of 100mA average current supplied as DC is 100mA. The RMS value of 100mA average current supplied as a 20% duty cycle pulse train (so 500mA peak) is 224mA. If the loss mechanism was Ohmic (I know it isn't in a thermionic diode, but indulge me for a moment) then the RMS voltage across the diode would also be raised by the same factor. So the average power being dissipated in the diode would be 2.24 x 2.24 = 5 times higher. I would call a factor of 5 'a difference'.


Correct, and so would I.

 
Posts: 6972
Joined: Sun Aug 14, 2011 11:02 am
Location: Co. Limerick

Re: Testing 5U4 rectifier

Post by Michael Watterson » Sat Mar 22, 2014 8:18 pm

Um the point of RMS is the waveform or duty cycle is irrelevant, that is already accounted for in doing a true RMS measurement. RMS volts (across the valve) x RMS current (though it)is the same watts as the DC Volts x DC current.

The linearity of R doesn't matter either as we compare REAL RMS. That's the whole point of RMS. It gives you the equivalent power to DC, the actual average Watts. Assuming there are no inductive or capacitive aspects.

The RMS value of 100mA average current supplied as a 20% duty cycle pulse train (so 500mA peak) is 224mA.

This means nothing to me. Either the RMS is 100mA or 224mA, it can't be both. Average current we don't care about, its meaningless. Average power though is RMS volts x RMS current.

 
Posts: 917
Joined: Sat May 26, 2012 5:41 pm

Re: Testing 5U4 rectifier

Post by Valvebloke » Sat Mar 22, 2014 10:15 pm

Michael Watterson wrote:
The RMS value of 100mA average current supplied as a 20% duty cycle pulse train (so 500mA peak) is 224mA.

This means nothing to me. Either the RMS is 100mA or 224mA, it can't be both. Average current we don't care about, its meaningless. Average power though is RMS volts x RMS current.


Perhaps I should have said:

'Consider a train of current pulses whose duty cycle is 20%. Let the average current in this train be 100mA. The peak current will then be 500mA and the RMS current will be 224mA.'

I think it would be almost impossible to misinterpret that.

Average current is not meaningless and while you may not care about it, I do. And so will anyone who is designing a power supply. And so did the people who drew up the valve specifications.

If Andy wants to know whether or not a valve meets spec, which I imagine is the reason he is doing the testing (the increasingly distant subject of this thread !), then he will measure its performance and check that against the datasheet. Here's the one for the 5U4G http://frank.pocnet.net/sheets/155/5/5U4G.pdf. At the bottom of the page is a series of curves showing what the valve should deliver. The x-axis is labelled 'DC output current'. This is equal to the average current through the rectifier (conservation of charge). Immediately above the graph is a line saying 'Rectified current - 225mA max'. That is also the average current. As far as testing against the spec is concerned it's average current which matters primarily. The RMS current only matters because, along with the heater power (obviously), it determines the valve's temperature, which may affect its performance.

VB

 
Posts: 706
Joined: Sat Jan 07, 2012 10:21 am
Location: Sleaford lincs

Re: Testing 5U4 rectifier

Post by Dr Wobble » Sun Mar 23, 2014 8:44 am

These are NOS Brimar . If you have any of these and want to sell them,I,d hold on because next week the way prices are going you'll probably be able to stick another 0 on the end of the price. I didn't buy them to deal in but to use,anyway I digress.

As noted on here, there is little info on how to test rectifiers with a valve tester. I havn't got a leakage tester but have been meaning to build one for a while. My homebrew tester has room in it for more circuitry, so it looks like I'd be better to build a rectifier tester, with a cap leakage tester incorporated in it. I,ll re read your posts today and see if I can come up with a circuit.

Thanks for your input,Andy.

 
Posts: 706
Joined: Sat Jan 07, 2012 10:21 am
Location: Sleaford lincs

Re: Testing 5U4 rectifier

Post by Dr Wobble » Sun Mar 23, 2014 12:30 pm

I've re read your posts and admit to getting quite lost at the mention of loss mechanisms and current pulses etc. All I really want to do is compare anode voltage against current delivered and perhaps test PIV.

As the typical 5U4G data sheet gives current compared to capacitor or choke filter without load, all I really need to do is wire it up to a transformer connected to a variac with a capacitor on the end. I'm putting it simply.

For the PIV test make a HV low current PSU and connect as per Michaels instructions.

Am I correct? Andy.

 
Posts: 706
Joined: Sat Jan 07, 2012 10:21 am
Location: Sleaford lincs

Re: Testing 5U4 rectifier

Post by Dr Wobble » Fri Aug 14, 2015 11:22 am

Revisiting this subject again, as I've about finished valve tester Mk11. To finish it, I need to bolt on a rectifier test. I'm a little further along IE I know a bit more than when I first asked the question.

I intend to test a rectifier with an isolated variac as the AC supply, however looking at the 5U4G data sheet I have a few question's.

It stipulates a 75 ohm supply impedance; no idea what my variac supplies Z is - will this matter?

Also on the data sheet at 200v RMS there is hardly any current. I'm confused! Is 250v mains supply P-P or RMS? If P-P how has the datasheet any relevance as the RMS of mains supply could be 170v ish. The other niggle I have is to why the datasheet shows curves up to 600v per anode. Most rectifiers like the 5U4G were made to regulate mains which is nowhere near 600v. I feel I'm missing something here, probably knowledge.

Lastly, as I want to keep things simple and test as near to how the valve would be connected as in an amp/set, I want to know 2 things about the rectifier under test. Does it work, how well does it work. So AC in via a resistor divider as per VB's post, with the center to earth *, HT off one side of HTR to a cap as load into a ammeter. This VT has taken ages to finish, I want it done and can always improve/add to it later.

* Two 100 ohms R's ?

Kick me up the arse metaphorically if I've written anything foolish.

Andy.

 
Posts: 4092
Joined: Wed May 02, 2012 3:35 pm
Location: Worksop

Re: Testing 5U4 rectifier

Post by Refugee » Fri Aug 14, 2015 12:14 pm

What you have there is a rectifier with a DC output of 275ma and at 275ma the forward voltage of one diode is 50 volts.
What I would do is to warm the heater up for a few minutes with an ammeter connected to cathode and ground and feed each anode with a variable smooth DC supply from a Variac and 50V RMS transformer with a smoothing cap and silicon diode and see what forward voltage drop you get at about 200ma. The maximum continuous current for one anode is about 130ma so take your measurements quickly to avoid red plating the anodes. The data sheet shows the voltage drop during one half cycle of mains and the other anode takes over for the next half cycle so it will give you time to take a reading with a DMM before anything gets too hot.
Once you are done with this and before you shut down the heater supply use a Megger to test with positive to cathode and negative to each anode and check the readings with the reverse leakage figures in the data sheet. It should read vary high on a 500V test.
http://www.r-type.org/pdfs/5u4gb.pdf


Return to General Work Shop Discussion



Who is online

Users browsing this forum: No registered users and 4 guests