The AM Forum

OK, following the other 'hv transformers' thread it was clear about how using a FW bridge with a transformer rated for a FW with C.T. results in the rated current needing to be divided by two since the DC output voltage doubles. This way the same V*A product exists with either rectification type.

But what about keeping the load current with FW bridge the same as FW-CT when the duty cycle is low? Now for short periods we deliver twice the rated power. The applications of powering a class B modulator, or SSB linear come to mind.

So, I got my 8877 shoe-horned into a Dentron MLA-2500 amplifier. With the stock power supply, I could easily do 300 watts AM, sometimes 400, and legal limit SSB. The power tranny got pretty hot though with AM. Now I need to recap, and instead I want to clear out the power supply and start over. The attached picture shows what I am contemplating. 

That Collins transformer rated at 2250VCT @ 0.565A. The secondary DC resistance is 90 ohms, and each primary 115v winding is 1 ohms DC. The current rating is CCS and at 10,000 feet. Any idea what the ICAS 50% duty cycle rating might be?

I figure with 230vac in, I will get about 3200V unloaded, and about 2750 loaded (~1.2*rms). For 375W AM linear at 33% efficiency, I need 1125 watts DC input, which could be 409ma @ 2750v. Is this doable?

For SSB unprocessed, I will let it rip....

Jim
WD5JKO

Here is something I posted a while back related to this topic.

Take a look at the link at the end. (but remember the guy was taking about linears for SSB).

Many of the old UTC CG plate transformers are designed to use with the center tap grounded -- you are warned about this issue in the catalogs. The reason is that the center tap is not insulated well enough to withstand the voltage that the center tap would see in a bridge circult.

In some cases, they have some versions of the transformer, that ARE designed to work in a bridge circuit -- this is indicted by a W as the final letter in the part number (e.g., a CG-303 vs. a CG-303W.

I belive that to run a transformer in a bridge circuit rather than a full wave circuit with the center tap grounded you can draw about 70% of the rated current rather than 1/2. In other words, if your transformer is rated at 1000v CT at 200ma that means you can get 500v at 200ma with a full wave circuit with the CT grounded, or 1000v at 140ma with a bridge circuit. (the output voltage depends on the type of filter used of course)

If you look at the old UTC catalogs, you will find some transformers are rated for either rectification circuit, and indeed, the current rating for the bridge circuit is 70% of the current for the grounded CT full wave circuit.

I posted an analysis a while back of this, but it went over like a lead balloon, but I still feel it's right.

The reason is that when you use a grounded CT circuit, only 1/2 the secondary is conducting at a time. When using the bridge circuit, the whole secondary is conducting all the time.

So if you run the transformer at the same output current in a bridge circuit you would have double the heat generated in the resistance loses in the secondary. But, if you decrease the current to 70%, then you have the same heat loses. The reason is that heat generated is I squared R. So you to get half the heating you need to derate by dividing the output current by the square root of two which is approximately the same as multiplying by 0.7.

Here is interesting link for info on using CCS transformers in various ICAS applications:

http://wb0nni.dakotamade.com/xfmrpwr.html

David,

    Ok on that post. Doing 70% rated current as you say would allow me to run AM at 375W out.

So I just put the tranny on the scale, and it's 25 pounds. The Orr chart then says:

.6KW CCS
1.2KW CW
1.4KW IVS
2KW+ IVS min

I find it interesting that the tranny says 200 Celsius on it.

Jim
WD5JKO

Be careful with what transformer model you are using. The commercial duty C  UTC version is OK to abuse as mentioned if it was designed for a FWB. The economy models are already running ICAS specs and others are sort of in between. The internet is full of old catalogs and this is one of the better sites. There are also plenty of documents on calculating the KVA by measuring the cores cross section if its not a sealed unit. In that case the weight minus a correction guesstimate for potting and sheetmetal has to be used; also they retain heat.

http://www.bunkerofdoom.com/

Old timers will remember the Ameritran brand that was popular surplus in the 50's. Those could take a real beating.

I ran a pair of 250TH's in the amp on a BC-610 transformer and 810 modulators on another. No problems but I didnt go into monologues on AM either. Mostly used the amp on CW chasing DX.

From what Ive read of the commercial BC TX wars in the industry magazines is that Collins transformers were considered light weights. Id suggest load testing before installing.

Carl
KM1H

Jim,

Sounds like an SB220 transformer.  Except for the 200c rating!

That's pretty warm!

--Shane

OK, well the transformer you have pictured is 2250 at .565 ma.

So assuming collins speced it like most transformer companies, that means 1125V at .565 for a FWCT circuit.

With a bridge, you would get .565 * .7 = .395 out of it.

So at 2250 you would get 2250 * .395 = 890 VA

Now assuming a linear with 375W output on AM, you would need a power supply that could produce 3 times that for full CCS operation.

That would be 1125 Watts.

And you have 890 so  you are 25% short.

BUT, it looks like that tranny you pictured has a Mil spec number on it. So it's likely conservatively speced.

Plus you aren't going to talk 24 x 7 on it, so the duty cycle would be 50% or so.

So it looks like it would work in a legal limit (375 W) linear for AM - barely.

It would probably work OK if you didn't deside to do a 1hr monologue. :-)

But it's none to heavy for the task.

I agree with Carl on my comments about UTC transformers. The comments I made apply to the CG and LS series power transformers. Those were commercial units. The S series where "amateur" and intermittant service and were speced at ICAS ratings.  

If you could jury rig up the supply and try it before cutting metal to install it, that might be a good idea.

Dave

Anybody have any information on the CG310. I read they were rated for 600 ma at 4KVDC CCS on the high tap.

CG310 specs:

4600-4050-3500 - 0 - 3500-4050-4600 at 600ma weight 150 lbs.

That's a big boy.

I have a CG308, but it's weighs a mere 125 lbs.

  Hmmm, I got all the stuff at my work. We have a nice power supply test station with a variac controlled 0-250 vac with V & I meters. To test the supply I need a dummy load, like a 5000-7000 ohm >1000 watt resistor. This makes me think of two probes in a large bucket of salt water. Anybody do this? Could use some advice on how to do that. Don't want to make a mess, or kill myself.  :(

Jim
WD5JKO

I was thinking more like hooking it up on a bread board and then attaching it to the amplifier and seeing how it runs on the air.

But it is dangerous to have such a high voltage supply bread boarded with lots of exposed connections.

Perhaps more trouble than it's worth.

Jim,

Search Calorimeter on google, or even better, there is an article on building one in the ARRL manual, circa 1995 that's sitting on the bookshelf.

--Shane

I regularly safely load test up to 7500V PS using resistive loads and a few times a big tube added in with a lot of air runing idle current only. All done on a wood covered raised basement floor, haywire style (no clip leads  ;D) but with proper insulation, etc. The load resistors also get lots of air blowing on them. Filtering is by one of the large oil caps I have on hand.

The 13A or 28A 240V Variac and meters are far enough away for safety.

Ive let them run for several hours to bake the iron or monitor temperature rise. No kids or pets allowed in the basement!

Carl

Last time I wanted to check some large iron I went from mains to the iron and used a second transformer back to mains. Put a few electric heaters on as a load.

The transformer under test was a modulation transformer with a 1:1 ratio. I ran 6kv through it and stuck 4kw worth of heaters on it. The thing didn't even get warm!

IIRC, Timmy did something like that years ago trying to send power to a remote location "up the hill" with ladder line. I seem to remember that the ladder line broke down and the whole mess went up in flames.

Dave, that must have been a STRAPPING mod transformer!!

THat's prolly the mod transformer Dave lugged up to Tom's place in the trunk of his Cadillac and it busted the concrete steps on Tom's porch  ;D

But it'll do a 500 watt input class c plate modulated rig halfway decent!

6 X 12 Uhaul. I backed into the porch while leaving.

The sign of REAL iron.  Shows up in a UHaul and causes damage while getting it into the house.

:)

--Shane

Jim,

Perhaps this comment is a bit off-topic, but will that 54uf cap shown in your accompanying photo be acceptable for use in this PSU you are considering? I ask this as perhaps that capacitor is designed for energy storage/discharge applications only, and it's ESR (equivalent series resistance) may not permit it's use as a filter cap due to the high ripple currents encountered in a typical HV DC power supply.

Just curious.

73,

Bruce

I have the same cap filtering my 4x4-125A amp, and it seems to be working fine so far. Doesn't get warm or anything.  No reports of hum on signal. Haven't checked with a scope for ripple voltage though...

Those storage caps work OK up to about 50-60% of voltage ratings in a ham style PS which is usually a doubler with lots of 60 Hz ripple. In a FWB they should run cool but always feel the case once in awhile until you get comfortable it will not run hot.

Carl

Carl,

Thanks for the information.

Are you suggesting that in a FWB (or possibly a FW CT configuration), the cap need not be derated 50 to 60%, and can run up to it's normal working voltage in an amateur-type PSU?

Perhaps you can clarify this for me, as I have one of these energy storage caps that I am considering using in a FW CT HV DC PSU for a 2KW class AB1 linear amplifier I am thinking of building.

Thanks & 73,

Bruce

some guy in Memphis has new Maxwell oil caps on eBay 5 KV 47 uF wants around $79 each, shipping another $25.
42 uF filter cap on full wave rectifier 4.5 KV DC give about 1.5% ripple.

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=330419830134

Yeah,  made with Ham Oil.

klc

We are having ham for Easter dinner. I'll save some of the oil if anyone wants it.

It might work really well in filter capacitors used in a pole-pig power supply.

If you look at the pix closely, the tag is blurry enough that you cannot read it. That cap looks awful small for 47uF @ 5K.  ???  ??? Also the terminal insulators look a little wimpy for that kind of voltage rating.

the 10s @ 4Kv in my 4X1 transmitter are the same size, if not larger. ???

Going by what Ive "heard" is that a FWB or CT will smoke them at rated voltage and be OK at around 70-75% due to the 120 hZ ripple. OTOH, I dont care to have any cap get much above room temp which should be your benchmark.

I have a pair of 53.5 uF @ 5KV discharge caps which Ive run here in a FWB in the past at 2400-2600V and they get moderately warm (very comfortable to the touch)after several hours of use (typically the PS were on 6-8 hours at a stretch) on a prior AM amp and modulator; 3 x 813 with 572B audio.



Would you trust a guy selling the item below??? >:( :o
290419274224

Carl

Saw this testing method for a plate power supply being tested at Universal Voltronics Corp:
15 kV at 6 Amps.
They had a big plywood sheet, standing upright, with over a hundred light sockets, the porcelan types. Screwed into
there were those little conical nichrome wire heating elements that used to be in every hardware store. All wired in series/parallel
to get the 2500 ohms needed, and then a bunch of window fans blowing across this thing. Simple, fairly cheap, but don't let
your cat walk next to it.

I had a photograph somewhere of this setup but can't find it today.

Update:

   My need to re-do this HV supply has taken a back seat since I fixed the original power supply in the Dentron MLA-2500. I cannot believe they (Dentron) used fiber insulators under the electrolytics all the way up the voltage ladder. The cap at the highest voltage had the metal can return about 1/10" away from the metal 6-32 screw. So the B+ (2500 - about 400v) had a 1/10" gap to a 6-32 screw head ground. It was arcing big time. I just replaced that screw with a 6-32 nylon screw. So I'm back up.

   It is good that others have used that same Aerovox 54 MFD 4400v capacitor with success. I called Aerovox and requested a data sheet. Still waiting, but if they send it, I will post it here.

    As far as that Tranny goes, I still think it is useful for this application. I never mentioned it, but I have two of them. If I put them in parallel, then I should have the basis of one heck of a brute supply. This supply will need to be in it's own chassis however which is probably best anyway.

   Thanks for all the suggestions.

Jim
WD5JKO

Frank, good point.  Yeah, I bet those are discharge caps.  I will use 3 GE 12 uF 5.2 KV oil caps (that's 5.2 KVDCW) along with the TOBE cap and get 46 uF 5.2 KV for the peak 4.5 kv from the transformer.  These all have large insulated threaded lugs; they are the real deal. those discharge caps would be okay in a 3 kv supply though.

To add to what Carl said about heating, realize that the heat will increase as the supply is actually delivering current.  So sitting at idle is not a sufficient test to see if the capacitors are going to get too hot so monitor them also during the first few hours of use in your application. 

Of course our class C finals running continuous carrier and class B modulators with a high resting current are going to stress the supply a lot more than one used for normal SSB usage.   So be careful in extrapolating web reports of performance over to the different environment of AM.

Yeah I wondered about that.  Gee, now I'm wondering if 700 v. margin (5200 - 4500) is enough.

Rob,

Are your GE caps energy storage or are they originally designed for filter work? 

Well, here they are:  http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=250536218736

I hope you're not gg to tell me they're no good.  I've blown over $400 on capacitors now  ???

Rob

Rob,

Those will probably be OK but in circuit testing will tell you for sure.  Do you have the actual GE part # from them?  The ebay picture wasn't detailed enough for me to see.  One of my colleagues came here after several years in GE engineering and he can probably get the product specs.

  I am a little confused about the two categories of capacitors mentioned on this thread, 'energy storage', and 'filter capacitors'.  Looking at some 'energy storage' capacitor data sheets, they have very high peak current capabilities, often in the thousands of amperes. This suggests to me that the ESR of these capacitors must be very low.
   I am thinking of the difference where the 'energy storage' cap is charged up, and then suddenly discharged into a near short circuit (thousands of amps into a single pulse) where the repetition rate is very low. Then compare this to a 60HZ power supply filter where the 'filter capacitor' might see an RMS current of  maybe an ampere where the current is periodic at a 120 hz rate.
   So wouldn't a low ESR 'energy storage' capacitor also make them suitable as a 'filter capacitor' whereas the reverse may not be true? Maybe someone can clarify the issue for the benefit of the group.

Jim
WD5JKO

There has been a lot of discussion of this on the Amps reflector at contesting.com.  There is a searchable list archive at:  http://lists.contesting.com/archives//html/Amps/

Like any place on the internet, there is a mix of fact and opinion and it is not always clear in which category any given post falls.  But you will find some good discussion and additional links on a range of topics from tubes and power supplies through esoteric topics like ampliphase modulation along with a lot of "refuse" on parasitics.

Well I had a really nice reply with everything I had to say and then I accidentally deleted it somehow.  In disgust over that I now offer an attempt at the same content but condensed:

I didn't know until recently that there were differences among oil caps.  It might help me and others if you guys who know how to get good caps for filtering in amp power supplies for AM service would say where you find yours for purchase. 

I've looked at a lot of oil h.v. caps on eBay and none of them say "filter cap."  They may say "energy discharge" which I now know means they are for photo flash units or science fair projects but usually you just get N uF and N KV with no indication the voltage is for working volts or peak volts.  This last time I went by size/uF and the big insulators on the lugs.  They also look like regular oil caps you'd see in a p.s. and not something goofy with spade lugs.  Of course I could just by new and pay $$$ or go to Henry Radio or Surplus Sales ($$$ there too).  What else is out there?  Hamfests do not seem to turn anything up, or else you find 1 uF at 10 KV or 50 uF at 500 v.  I got these three 12 uF 5.2 KV caps for around $160  and figure paralleled with my 10 uF 8 KV cap will get me 46 uF 5.2 KV for a 4.5 KV DC supply.  I guess when I finally fire that baby up with a pair 4x1 I'll check them after 30 minutes operating and see if they're hot but if you guys could give some pointers to sources for good used filter caps other than ePay I'd appreciate it.    tnx

Rob

Rob,

Old vintage oil caps are going to be standard filter service caps but some of them will contain PCB if that concerns you.  Hamfests and parted broadcast rigs are both good sources for these older caps.  I used new GE caps in my HB amp (8 times 7.1 uf @ 5 KV) but they are only exposed to 2,200 volts.  They have been in use for heavy CW and SSB contest use for about 8 years without failure and with no detectable temperature rise but I have no data on what happens when the input voltage approaches the nameplate rating.

With the energy storage caps you either need to get the manufacturers specs or test them under load and check temperature rise.  There are a lot of opinions out there but they are generally based on one ham's experience with one or two types of caps and extrapolating that experience out to "energy storage" or "pulse" caps in particular would be dangerous.    For example High Energy Corp. shows that ripple voltage needs to be reduced as frequency increases with a maximum of 20% at 60 hz and dropping to 10% at 400 hz indicating that with a full wave bridge (120 hz ripple) some de-rating is needed.

Okay Rodger tnx, i'll use these and see what happens.   Broadcast rigs (or at least the one I have) come up short in the cap department.  AM tube tx used filter chokes b/c regulation pertained more to constant current with v. swings so the caps are fewer and relatively low v.  I think the most I got out of my bc tx was 4 kv and 4, or maybe 8 uF.    These ones I'm getting don't have PCB.  Now, the one that's already in there, the 10 uF @ 8 KV... ::)  ;D  Okay you are running low v. with a lot of paralleled caps--I see; well, Handbook says a few hundred v. over the peak DC p.s. v. should be okay.  Anyway, we'll see.  I think they will probably be okay.   8)

Rob

As I alluded to earlier in this thread.

Dont put all discharage type capacitors in one bucket either, they vary considerably depending upon application. Its a lot like comparing peak power out of a radar tube; it all depends upon the PRF and pulse width. You can get hundreds of KW out of a tube that fits in your palm if the duty cycle is very low. If something goes wrong then it become PIR....Premature Involuntary Retirement ::) ;D

A filter cap is designed with heavy foil and insulation material plus cooling oil and takes up a lot of volume per uF. Discharge caps get their high uF by reducing all of the above.

Since there are so many varieties of discharge caps, ratings and terminology vary by manufacturer and application, then the only real test for filter use is reports by others and the heat test. If it runs cool its OK. If it runs quite warm then you take the Dirty Harry quiz 8) :o

Carl
KM1H

Premature Involuntary Retirement...that's a good one hi hi.   Another thing I've learned to not do is jam the caps all up against each other; instead leave some room between each one to let air circulate around them.  That was a great tip-I have seen them packed in and never thought about heat until now.

Rob

Sounds like voltage doubler service would really ask for trouble.

dupe, deleted

Buy using the same logic in reverse, one would think motor-run AC capacitors would be particularly good when used as filters in a HV supply.

They are designed for AC -- all ripple all the time.

All ripple until the motor gets up to speed. Usually pretty fast.

that's the motor-start cap, not the motor-run cap?  ;D

Rob,

I looked through my Radio Masters catalogs (I have most of them from the late 50's through early 70's) but they were not of much use.  Although the catalogs are very large there were not a lot of listings for oil filled capacitors and the GE advertisements did not provide much information.  I did find that in addition to energy storage caps GE also made another group of oil filled caps that were designed for light duty coupling and buffering usage where the ripple is low and are known as the CP-70 series so make sure yours isn't in that model series.

On a side note, these catalogs are a generally great reference source (they run from 1,000-2,000 pages in length) and provided detailed specs on a lot of components.  If you see them hanging around at a hamfest they are worth picking up.  Mine came as a free bonus with my $75 Mosley CM-1 receiver and I have used them more than the receiver.

Im under the impression that the motor run cap is mostly along for the ride unless it needs to provide a shot to maintaining the 90* phase relation with the current. Even then its sole duty is voltage with the windings providing the current. Sort of like a PEP PS but not quite....see below..

In many motors the same cap is used for starting except in high torque applications where a seperate cap with higher current ratings is used and now we are approaching a real PS.

Motors can be more confusing than antennas :-X

Carl
KM1H

true, I don't know much about them... the one-cap motors I am familiar with have a centrifugal switch to cut the start cap out of the circuit once the motor is up to speed... I know there are two-cap motors, one for start and one for run though.

Rodger tnx vy much.  you didn't have to do that but I appeciate hearing about Radio Masters.  I looked at the Sangamo 8 uF 4 KV cap that came out of the broadcast rig p.s. and it is a bit smaller than the GE caps I ordered, but about in proportion for the value and voltage with the GEs and similar construction on the outside (i.e. the insulators) if that means anything, probably not much.  I'm pretty confident they'll be okay for testing and as someone, Carl I think wrote, the surefire test is to fire up the rig and run it for awhile and see if the caps are hot. 

Meanwhile, tonight I put a 5 v. drop transformer in an outlet box with a cord going to it to step the 120 v. line down to 115.  I want to free up a variac for more useful work.  I test it and read 118 v. and find my service v. has gone up 3 v. from 120 to 123.  Can't win.   ???

Rob

Rob,

You are welcome.  From what I could see in Radio Masters your caps are considerably larger than the energy storage caps and the only wild card are the GE caps that are neither listed for energy storage nor filter use.  Too bad they did not provide more detail but I will try to run down some info.  I will scan and post that page from Radio Masters for info purposes.

As to your line voltage, don't worry it will go back down again with the summer air conditioning load.  You really can't win :)

With all of the discussion of the difference in application for "energy storage" caps vx "filter type" caps. I wonder what is actually different in the physical construction of the capacitors?  ???  ???

I answered that yesterday.

Carl

the difference seems similar to lead acid batteries; deep cycle versus normal car batteries. The deep cycle will be larger, and have thicker plates whereas the normal car battery has very thin and fragile lead plates.

Thanks,
The line v. is usually dead on 120 v.  I wish it were lower but I have not seen it above that before.  I guess modern electrical loads can deal with input v. swings.  I pluged the 75A3 into the step down box and fired it up.  It made a little noise I had not heard before and I lifted the top lid and looked inside.  It sounded different and I didn't like that so I turned it off.  THEN I rechecked the line in the shack which of course, I should have done first.  It was 123!  coming out of the box, 118.  I'm gg to have to simply buy a bunch of variacs, put them in boxes with v. meters.

There are several problems with eBay, one being poor item descriptions.  I should develop the habit of asking questions on there.  Two other problems are sellers selling stuff they know nothing about, sellers ignoring categories (transmitters listed with receivers for example) and lack of subcategories.  There are around 20,000 listings for capacitors.  There should be sub-categories for air variables, micas, ceramics, oil, ...the same goes for other parts.  There are other problems but that is enough as I have already veered way off topic.

Rob

that would help... but just searching for "high voltage capacitors" shows a fair number of interesting (if expensive) ones, especially the Russian mil-surplus  :D

I wonder how putting ten 560 uf, 450v electrolytics in series (with equalizing/bleeder resistors) would work instead?

see reply 6 7 at:
http://amfone.net/Amforum/index.php?topic=17729.msg122355#msg122355

Update, I got a reply from Aerovox concerning the 54 MFD @ 4400 vdc capacitor pictured at the beginning of this thread. So with an ESR of 0.2 ohms, I don't see a lot of heat from a amp or two of RMS ripple since for 2 amps, W=I^2*R, or 2^2*0.2 = 0.8 watts.  Take 800mw, and oil cool it in a cap the size of a 12oz beer can, and we should not get hot. I do see how a de-rate of voltage might be prudent. In my case, 3KV with a 4400 v rating puts it at 68% of spec. I think that would be fine. Any thoughts?

I pasted the Aerovox reply below.

Jim
WD5JKO


From: eperry AT aerovox DOT com
Sent: Wednesday, April 07, 2010 10:22 AM
To: Candela, Jim
Subject: Fw: Need Capacitor Spec Sheet
 Jim

Thank You for your inquiry.  The item listed below is obsolete.  Per our engineering dept this was a light duty capacitor and do not have  an RMS current rating.
The ESR would be 0.2 ohms.
If you have any technical questions you can contact the Engineer Paul Winsor 508-910-3178

Thank You
Ester Perry
Account Manager
508-910-3666

My thoughts would be that it would probably work fine at 3KV. As for the ripple, run it a while and if it doesn't get hot, it's probably good to go.

Be careful about putting your hand on it to check the temp while it is powered up.

I have one of the gun type laser IR meters and if you have one of those you could take the temp while it is running without touching it.

It might be interesting to call that Winsor fellow and see what he says about voltage derating and using it for a PS filter.

I'd be interested too... makes "Working Voltage" rather meaningless if the cap won't work at said voltage  ::)

Meanwhile I am running a similar 56 uf, 4.2 kv cap at 3.8 kv standby, 3.6 kv idling current, will see how it holds up with my infrequent usage  8)

The operative phrase from Aerovox is  this was a light duty capacitor and do not have  an RMS current rating. .

Try the 68% and keep monitoring it as heating damage tends to be cumulative. I wont risk stepping on my crank and saying it will be fine but you have decent odds in your favor ;D

Carl