[Amps] .01uf caps across diodes ?

Wed Dec 14 03:49:27 EST 2022

During the early 1960s I worked at a BC station where we had a 5 kW FM 
transmitter. It had a full-wave bridge rectifier consisting of 4 boards, 
each of which had some 16 400-piv diodes on it. I think (but don't 
remember for sure) that DC plate voltage was about 3200. I think the 
boards had a resistor across each diode.
Every once in a while there would be a sound like an M16 on full 
automatic as the diodes on one of the boards would let go. Replacement 
boards came with capacitors, which I believe solved the problem.

73,
Victor, 4X6GP
Rehovot, Israel
Formerly K2VCO
CWops no. 5
http://www.qsl.net/k2vco/

On 14/12/2022 9:55, sm0aom--- via Amps wrote:
> The large characteristic spread of silicon rectifiers of "yesteryear" often resulted in reliability problems.
> 
> One early task I got as a systems engineer in 1981 was to investigate why the 6 kV six-pulse bridge rectifiers in a total of 15 5 kW MF/HF SSB transmitters failed randomly. The reasons were found to be transients on the mains supplies, caused by switching of inductive loads.
> 
> The safety factors in the expensive rectifier "sticks" originally used were not that large, and a contact with an applications engineer at German semiconductor manufacturer Semikron netted a booklet about the benefits of controlled avalanche characteristics, as they handle transients much better. No parallel equalising networks also were needed.
> 
> The rated PIV of the originally supplied rectifier sticks was only 8 kV, so a design goal of 12 kV was chosen. 2 A DC load was used as a rated current into the choke-input filter.
> 
> Semikron had a controlled avalanche diode, SKa 1/17 which permitted an average current of 1.2 A at 70 C ambient rating, and 1700 V PIV. 8 diodes in each leg provided somewhat additional safety factors.
> 
> It turned out that the rectifier reliability problems vanished entirely after replacing the old "sticks" for the whole remaining service lives of the transmitters, which finally were scrapped in the mid-90s for other reasons.
> 
> 73/Karl-Arne
> SM0AOM
> 
>    
> 
> ----Ursprungligt meddelande----
> Från : 4cx250b at miamioh.edu
> Datum : 2022-12-13 - 18:21 ()
> Till : g8gsq at gmx.com
> Kopia : amps at contesting.com
> Ämne : Re: [Amps] .01uf caps across diodes ?
> 
> An interesting point, Steve. If I understand you, you’re saying that
> if a modern diode in a series string approaches reverse breakdown, it
> will start to conduct, thus shifting its overload voltage to other
> diodes in the string. Before the advent of controlled avalanche
> diodes, however, an overloaded diode would just short circuit, thus
> permanently shifting its voltage to the remaining diodes in the
> string. This process, once started, would likely lead to the
> destruction of the entire string.  To me, the takeaway message is that
> there’s little to be gained these days by adding .01uF caps to each
> diode. This is a helpful explanation, Steve. Thanks!
> 73,
> Jim W8ZR
> Sent from my iPhone
> 
>> On Dec 13, 2022, at 1:28 AM, Steve Thompson <g8gsq at gmx.com> wrote:
>>
>> It's not only uniformity, also the 'controlled avalanche' of modern diodes. As you reach breakdown they don't fail instantly from too much reverse voltage, they start to conduct slightly. In a string of diodes this results in voltage sharing happening automatically.
>>
>> Steve G8GSQ
>>
>>> On 13/12/2022 02:11, MU 4CX250B wrote:
>>> Hi Bill, I remember that argument,too,  but It doesn’t carry much
>>> water with today’s uniform components. The back resistance of a series
>>> string of diodes is so high that I doubt an uneven capacitive voltage
>>> divider would cause problems. Normally one designs an HV diode bank
>>> with at least a breakdown voltage safety factor of 2, and preferably
>>> more.
>>> 73,
>>> Jim W8ZR
>>> Sent from my iPhone
>>>>> On Dec 12, 2022, at 10:19 AM, hzp_electronics at juno.com wrote:
>>>> Hi Jim
>>>> The best argument that I have heard about capacitors across diodes is that older diodes had uneven capacitance.  Under certain conditions the capacitance would cause the voltage to divide unevenly between the diodes.  The added capacitors were made large enough to swamp out the capacitance difference and keep the voltage division more equally.
>>>> \
>>>> Whether the caps are needed with modern diodes, depends!
>>>>      73
>>>> .... Bill K3HZP
>>>>
>>>>
>>>> ---------- Original Message ----------
>>>> From: MU 4CX250B <4cx250b at miamioh.edu>
>>>> To: Tim Duffy <k3lr at k3lr.com>
>>>> Cc: jim.thom at telus.net, amps at contesting.com
>>>> Subject: Re: [Amps] .01uf caps across diodes ?
>>>> Date: Sun, 11 Dec 2022 22:32:29 -0800
>>>>
>>>> Hi Tim,
>>>> I’ve followed this debate for years and frankly I don’t think adding
>>>> .01uF caps across each diode in a diode string makes much difference
>>>> one way or the other. It’s true that there will be some suppression of
>>>> fast current and voltage transients, but these aren’t likely to cause
>>>> damage to modern components. To me the downside of the practice is
>>>> that adding 0.01uF caps results in additional opportunities for
>>>> capacitor failures, loose solder connections, unsightly wiring, and a
>>>> bunch of extra components that provide minimal benefits. I’m inclined
>>>> to leave them out. (BTW, one commenter suggested that sharp voltage or
>>>> current spikes wouldn’t make it through the filter capacitors. That’s
>>>> untrue because the filter caps have a self inductance that keeps them
>>>> from filtering spikes with high frequency components. )
>>>> 73,
>>>> Jim W8ZR
>>>>
>>>> Sent from my iPhone
>>>>
>>>>> On Dec 11, 2022, at 7:03 PM, Tim Duffy <k3lr at k3lr.com> wrote:
>>>>>
>>>>> W8ZR Jim
>>>>>
>>>>> Can you give advice on this topic?
>>>>>
>>>>> 73
>>>>> Tim K3LR
>>>>>
>>>>> -----Original Message-----
>>>>> From: Amps [mailto:amps-bounces at contesting.com] On Behalf Of jim.thom
>>>>> jim.thom at telus.net
>>>>> Sent: Tuesday, November 29, 2022 2:28 PM
>>>>> To: amps at contesting.com
>>>>> Subject: [Amps] .01uf caps across diodes ?
>>>>>
>>>>> What's the consensus on wiring .01uf disc caps across diodes in a HV supply
>>>>> ?
>>>>> W8JI sez they are required, and used in ameritron amps....and used to
>>>>> eliminate switching transients.
>>>>>
>>>>> Some say the caps are required to meet emc specs on commercial gear.
>>>>>
>>>>> I can't see how anything can get past the HV fitter caps.
>>>>>
>>>>> KM1H also uses caps across diodes.  He also used another method, with a
>>>>> .01uf cap across an entire series string of diodes, like a .01uf @ 5 kv
>>>>> type.
>>>>>
>>>>> Orr's books recommend caps across diodes... to eliminate noise on TX on
>>>>> ssb. .
>>>>>
>>>>> Are they required...or not ?
>>>>>
>>>>> Jim  VE7RF
>>>>> _______________________________________________
>>>>> Amps mailing list
>>>>> Amps at contesting.com
>>>>> http://lists.contesting.com/mailman/listinfo/amps
>>>> _______________________________________________
>>>> Amps mailing list
>>>> Amps at contesting.com
>>>> http://lists.contesting.com/mailman/listinfo/amps
>>> _______________________________________________
>>> Amps mailing list
>>> Amps at contesting.com
>>> http://lists.contesting.com/mailman/listinfo/amps
>> _______________________________________________
>> Amps mailing list
>> Amps at contesting.com
>> http://lists.contesting.com/mailman/listinfo/amps
> _______________________________________________
> Amps mailing list
> Amps at contesting.com
> http://lists.contesting.com/mailman/listinfo/amps
> 
> _______________________________________________
> Amps mailing list
> Amps at contesting.com
> http://lists.contesting.com/mailman/listinfo/amps