|
Components and Circuits For discussions about component types, alternatives and availability, circuit configurations and modifications etc. Discussions here should be of a general nature and not about specific sets. |
|
Thread Tools |
11th Sep 2004, 12:01 pm | #1 |
Nonode
Join Date: Jun 2004
Location: Norwich, Norfolk, UK.
Posts: 2,543
|
Reservoir/smoothing cap voltage question
Hi,
I have a valved variable regulated 0 to 300V HT power supply. I am checking it over and have found that one of the sections of the reservoir/smoothing cap will not reform. The sections are 8 and 16uf and the voltage rating is 500 VDC. I have modern 10 and 22uf replacements but unfortunately the voltage rating is only 450VDC. Is this close enough to use or should I find some with the correct voltage? Thanks David
__________________
http://www.youtube.com/ My Nixie Clocks |
11th Sep 2004, 12:33 pm | #2 |
Administrator
Join Date: Dec 2002
Location: Cardiff
Posts: 9,060
|
Re: Reservoir/smoothing cap voltage questio
You need to know what voltage they will have to work at. The ones on the output obviously only have to work at up to 300V, so 350V or higher will be fine here.
I'd be surprised if the input side was working at much more than 350V or so, as it seems pointless regulating a much higher voltage down, so chances are that 450V caps will be fine. I'd fit a 450V cap, and measure the voltage across it with no load on the output of the unit. If the voltage is less than about 420V, your 450V cap will be fine. A new 450V cap will be fine running at up to 500V for a little while - plenty long enough for this test. |
11th Sep 2004, 1:53 pm | #3 |
Retired Dormant Member
Join Date: Mar 2003
Location: Birmingham, West Midlands, UK.
Posts: 785
|
Re: Reservoir/smoothing cap voltage questio
...If you have any doubt, you could always use two 22uf caps in series with bleeder resistors across them to equalise the voltage (470k - 1 meg across each should do). This would give you 11uf at 900v - should be more than adequate.
|
11th Sep 2004, 6:45 pm | #4 |
Guest
Posts: n/a
|
Re: Reservoir/smoothing cap voltage questio
Remember to be very careful if your supply has a directly heated rectifier. It's best to check these voltages by measurement, then add at least a 10% margin for mains variations. Two capacitors in series works well as long as you use bleeder resistors as mentioned. Watch the VOLTAGE rating of any resistors you use.
Leon. |
11th Sep 2004, 7:39 pm | #5 |
Nonode
Join Date: Jun 2004
Location: Norwich, Norfolk, UK.
Posts: 2,543
|
Re: Reservoir/smoothing cap voltage questio
Thanks guys for the advise. Quantum's idea of using caps in series worries me so I think I will go with Pauls idea but I will use my trusty series lamp limiter to start with to keep the voltage down a bit
As for Leon's post it has a selenium rectifier, the valves are EL34, EF86 and a GD 150 M ? To me the thing seems much too complicated for what it does, so I must be extra carefull! David
__________________
http://www.youtube.com/ My Nixie Clocks |
11th Sep 2004, 8:43 pm | #6 |
Retired Dormant Member
Join Date: Mar 2003
Location: Bishop's Waltham, Hants, UK.
Posts: 939
|
Re: Reservoir/smoothing cap voltage questio
be careful with the voltage rating, as with a selenium rec, the smoothing caps will see the peak voltage across them until the valves warm up, also the EL34 is a fairly high impedance device, so I would't be surprised if its anode sits at nearer 400V. I suspect that 500V may be a safe minimum.
To check, measure the voltage at the anode of the EL34 at switch on, preferably with a high impedance meter. It is best to allow about 30% on top of this to allow for transients. Jim. |
19th Sep 2004, 11:02 pm | #7 |
Nonode
Join Date: Jun 2004
Location: Norwich, Norfolk, UK.
Posts: 2,543
|
Re: Reservoir/smoothing cap voltage questio
I've finally got around to checking the voltages. I connected a DVM across the 8uf 450V cap and applied the mains. The voltage was 508V ! there was a short but loud buzz at which point I switched off and ran away
Can anybody explain why there is soo much voltage inside when the output is only 300V and what is that alarming noise that it makes Thanks David
__________________
http://www.youtube.com/ My Nixie Clocks |
20th Sep 2004, 12:37 am | #8 |
Retired Dormant Member
Join Date: Jan 2003
Location: Livermore, California, USA.
Posts: 139
|
Re: Reservoir/smoothing cap voltage questio
Hi David
When you don't have a load on your power supply voltage before the regulator will be very high. When you start drawing current from the supply this voltage will come down. You need a cap rated for highest voltage conditions. Norm |
20th Sep 2004, 8:38 am | #9 |
Retired Dormant Member
Join Date: Dec 2003
Location: North London, UK.
Posts: 6,168
|
Re: Reservoir/smoothing cap voltage questio
A bit more detail. This is not intended to be a full explanation about how rectifiers work!
If you connect a smooting cap to a rectifier and there is no load the cap will charge to the PEAK voltage of the sine wave. This is SQRT(2) times the RMS voltage, eg 325V for nominal 230V mains. As you connect a load the mean voltage across the cap will fall but the peak voltage will not fall by much. In a regulated power supply the minimum (trough) voltage across the cap must be greater than the required output voltage. This is to allow the regulator something to work with. So a PSU with 300V regulated output at full load might easily have a waveform on the cap swinging from 310V up to 450V. And even higher on light load. |
20th Sep 2004, 7:19 pm | #10 |
Retired Dormant Member
Join Date: Mar 2003
Location: Birmingham, West Midlands, UK.
Posts: 785
|
Re: Reservoir/smoothing cap voltage questio
David,
Still not interested in using two caps in series? Seriously, it is a perfectly acceptable way of increasing the voltage rating of caps - most good radio books will discuss it along with the formula for working out capacitance of different value caps in series. Incidentally, your GD150M is a voltage stabiliser valve, I suspect that the 150 is it's mean voltage handling. |
20th Sep 2004, 7:56 pm | #11 |
Nonode
Join Date: Jun 2004
Location: Norwich, Norfolk, UK.
Posts: 2,543
|
Re: Reservoir/smoothing cap voltage questio
Hi Quantum,
I must admit your suggestion does look like the best option. I have heard bad things about caps in series but I had'nt heard of using " bleeder " resistors before to equalize the voltage. I've been doing some research into this and I've found this excellent page http://www.martinos.org/~reese/electrolytics/ I intend to use two 32uf caps to make the 16uf and two 22uf caps in series are going to be close enough for the 8uf. I have said before the circuit seems too complicated for what it has to do I am now not even sure that these caps are for resovoir/smoothing! I think they are for something else? Anyhow can anybody tell me what the loud buzz is? It's this sort of sound an ark welder makes when the welding rod sticks? Thanks David
__________________
http://www.youtube.com/ My Nixie Clocks |
20th Sep 2004, 8:30 pm | #12 |
Retired Dormant Member
Join Date: Dec 2003
Location: North London, UK.
Posts: 6,168
|
Re: Reservoir/smoothing cap voltage questio
The GD150M is a voltage stabiliser tube. It will try to maintain 150V across it regardless of the current flowing through it. It's not perfect so the voltage will vary slightly with current, temperature etc. It is normally used as the reference voltage for the stabiliser.
A fraction of the output voltage is compared with the reference and if the output is too low the main regulator valve is turned on a bit more (assuming it's a series regulator as I would expect) and vice versa. It's really a sort of amplifier with negative feedback. |
20th Sep 2004, 9:10 pm | #13 |
Dekatron
Join Date: Oct 2003
Location: Croydon, Surrey, UK.
Posts: 7,549
|
Re: Reservoir/smoothing cap voltage questio
Hi David.
Quantum is dead right about electrolytics in series . I've done it many times when adequate voltage ratings and/or values are not available. I once needed a 8uF at 500v in a Ferguson radio. It had a directly heated rectifier (which was new) and the peak voltage at switch-on was near to 480v. So I used a couple of easily obtainable 16uF 250v caps in series with a 100K 1 watt resistor across each. I did the same thing with the smoothing cap but used two 32uF 250v to give me the required 16uF. All of these fitted nicely inside the original smoothing can so the repair was disguised. I could have used 350v types but these would have had to be fitted under the chassis and the set would have looked less original. The value of the parallel resistor is not too critical although I would not go lower than 100K. Rich.
__________________
There are lots of brilliant keyboard players and then there is Rick Wakeman..... |
20th Sep 2004, 9:43 pm | #14 |
Retired Dormant Member
Join Date: Nov 2003
Location: Cape Cod, Massachusetts
Posts: 59
|
Re: Reservoir/smoothing cap voltage questio
Tim Reese's page cited above is an excellent discussion of the topic, and I hate to disagree (I see him from time to time), but the equalizing resistors are not usually necessary, provided the two capacitors are from the same production lot, and you don't try to run them close to their maximum working voltage. The voltages are self-equalizing to a large degree. And it's easy to forget that resistors themselves have a voltage rating, often not more than 250V, and adding them could actually cause problems.
When you buy a manufactured 600V electrolytic, there are no resistors inside, just two series caps. |
20th Sep 2004, 9:59 pm | #15 | |
Nonode
Join Date: Jun 2004
Location: Norwich, Norfolk, UK.
Posts: 2,543
|
Re: Reservoir/smoothing cap voltage questio
Quote:
David Last edited by Paul Stenning; 26th Dec 2004 at 8:35 pm. Reason: Fix link or code for vBulletin |
|
20th Sep 2004, 10:31 pm | #16 |
Dekatron
Join Date: Oct 2003
Location: Croydon, Surrey, UK.
Posts: 7,549
|
Re: Reservoir/smoothing cap voltage questio
Yep!!
I agree with Alan about the parallel resistors. However I was using capacitors that I had to hand and although they were of the same type, they were definately not of the same batch or even the same year of manuafacture. I tend to use resistors anyway because a) they are cheap (even the 500v ones) and b), they help to discharge the capacitors more quickly making the set safer to work on. Also in the case of the Fergy set on my previous posting, the peak voltage was about 480v which was fairly close to the maximum combined voltage. Rich.
__________________
There are lots of brilliant keyboard players and then there is Rick Wakeman..... |
20th Sep 2004, 10:35 pm | #17 |
Guest
Posts: n/a
|
Re: Reservoir/smoothing cap voltage questio
It's difficult to form aluminium oxide to give a dielectric strength much over 500V. In theory, balancing resistors are not required if the two capacitors are matched to the extent that the one of lesser capacity is not overrun on voltage (Q = CV).
However, electrolytics are by no means perfect capacitors, and have an appreciable leakage current. This is not necessarily matched between two seemingly identical capacitors in series, and any appreciable difference will result in a significant inbalance of voltage. By using a shunt potential divider, an artificial leakage of about 10x the actual value is introduced to " swamp " the difference. On high voltage equipment, I'd also be very glad of the self-discharge thereby introduced! Leon. |
21st Sep 2004, 11:37 am | #18 | |
Retired Dormant Member
Join Date: Nov 2003
Location: Cape Cod, Massachusetts
Posts: 59
|
Re: Reservoir/smoothing cap voltage questio
Quote:
Regarding swamping the leakage with resistance, it's worth noting that modern electrolytics have extremely low leakage, on the order of 10 or 20 microamps. This is vastly better than those made 30 years ago. How they will hold up with time, I couldn't say, but I'm optimistic. Last edited by Paul Stenning; 26th Dec 2004 at 8:36 pm. Reason: Fix link or code for vBulletin |
|