Forums :: Resources :: Features :: Photo Gallery :: Vintage Radio Shows :: Archives :: Books
Support This Site: Contributors :: Advertise


It is currently Sep Sun 24, 2017 6:21 pm


All times are UTC [ DST ]





Post New Topic Post Reply  [ 78 posts ]  Go to page Previous  1, 2, 3, 4  Next
Author Message
 Post subject: Re: Best way to turn 125 VAC to 110 VDC
PostPosted: Sep Thu 14, 2017 12:15 am 
Member

Joined: Jun Fri 19, 2009 6:34 pm
Posts: 6970
Location: Long Island
Quote:
You suggested that charging a couple of caps at a few hundred uF each directly from the line would blow the circuit and launch tubes into the air, as well as being dangerous to the person working on it (not hyperbole?). Lots of TVs charge up two large value filter caps in voltage doublers, which also draw fairly high current on the input side like we have here.


No, I suggested that charging two to ten thousand uF's of capacitors off a bridge rectifier connected to the AC line would likely trip the house breaker, but if it didn't, then discharging them through the radio might launch tubes into the air figuratively if not literally. I still maintain that would be the likely outcome of such an approach which fortunately appears to be the one not taken.

Quote:
If you read what has already been written, you'll see that we have been discussing two stage filtering all along, with a big resistor exactly as needed. Surge limiting also included from very early on in this thread. It would work fine. No 30 amp rectifiers needed. No circuits blown. No tubes launched. No huge (i.e., thousands of uF) caps needed.


I do not deny that you have been talking about a classic power supply with a two section filter and surge limiting. But taking shots in the dark on the values of a couple of parts to try is only talk, not a design. You are implying that with common parts like a rectifier, a couple hundred uF's worth of capacitors, and a resistor or two, you can deliver both sufficient voltage regulation not to endanger the tubes in the radio and at the same time deliver the needed output voltage while filtering out enough ripple for the radio to play as it was meant to. Well, where's your math?

_________________
"Hell, there are no rules here--we're trying to accomplish something!"

Thomas A. Edison


Top
 Profile  
 
 Post subject: Re: Best way to turn 125 VAC to 110 VDC
PostPosted: Sep Thu 14, 2017 1:02 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 11573
Location: San Jose, CA USA
The circuit is not particularly complicated, so with some simple Ohm's Law calculations and some scaling from known successful designs, a reasonable design has been proposed.

Basically, 10 ohms current limiting resistor on the input, followed by a bridge rectifier with diodes rated for an average current of a few amps each, followed by a pi filter of two capacitors of several hundred uF each, with a voltage dropping resistor between them. Exact current is not known (I estimate 500 mA based on actual design of radio), so start with 100 ohms and adjust up or down (most likely down) from there. Power dissipation in resistor is large, at 20 watts plus, so the benefit (to reduce resistor dissipation) and cost (loss of filtering with a smaller resistor) of using a bucking transformer was mentioned.

Cap size based on scaling by current ratio compared to known successful designs.

Failure of the diodes in the original attempt, provided a current limiting resistor was installed as recommended, was probably warm-up current of tubes, which could have pushed the average current well over 1 amp for 10-20 seconds. Probably enough to burn out the 1 amp diodes. Using diodes rated at 5 amps should do the trick, probably even less will do, since diodes are rated for average half-cycle current, and higher peak currents are typical in all kinds of rectifier circuits.

Neither you nor I knows exactly how much ripple this radio can tolerate without making too much hum, so I mentioned that a regulator might in fact be needed simply to suppress hum. From a voltage regulation point of view, however, it is not needed. Load, which is mainly tube filaments and speaker field coil, will be quite constant after warm up. During warm up, current will be high and voltage will be low. This will cause no harm whatsoever, other than a somewhat slow warm up.

I'm perfectly comfortable with doing more detailed calculations, but usually it's best to start with a clear picture of how it should work, and reasonable estimates.

It's obvious there is no good reason for this thing to blow breakers, etc. A radio like this doesn't need a lot of power. The mere fact of converting from AC supply to DC supply isn't going to change this radio from something normal into a beast that blows the breaker, unless poor design choices are made.

_________________
Tom K6VL


Top
 Profile  
 
 Post subject: Re: Best way to turn 125 VAC to 110 VDC
PostPosted: Sep Thu 14, 2017 2:17 am 
Member

Joined: Sep Thu 23, 2010 6:37 am
Posts: 9084
Location: Powell River BC Canada
"A couple of things I forgot to mention. My sig gen doesn't go down to 130 kHz IF, so I aligned it by ear. Seems good. When I inject 400 Hz into the 2AF tube, the output is clear. When I inject it into the 1AF tube, it seems the distortion is noticeably greater. Thanks,
Dave"

The 2AF tube grid is connected to a transformer secondary. The winding is shunted by a yellow
and red, 2nd AF grid resistor. I would suggest testing it to see if is open.
Resistor open or not, the 400 H Z from your generator is a level high enough, to inject
a clear signal onto the 33 tube. Even if the generator had no coupling capacitor, there is little
to interfere with the tube bias. The bias is established by the filament series feed system.

Moving back to 1AF, that tube has gain, and the audio input transformer, multiplies your 400
Hz signal, perhaps greater than the expected level from the 2nd detector. That alone could cause
distortion. But unlike 2AF, if your generator has any DC resistance, the bias developed by the red and
yellow, 1st Audio grid leak will be killed dead.

So the tack to take, is to make sure there is a capacitor in series with the hot lead of the 400 Hz.
output of your generator. Then see that you can attenuate that signal, low enough,
to inject just enough audio into 1AF grid, to produce the same level you got at 2AF grid,
(with no attenuation added)

There is an issue about where the common (ground lead) of the 400 Hz generator is to be placed.
The best place is across the respective yellow and red grid resistors.
Such a connection , sets up a shock hazard, and the possibility of blowing tubes if
accidental slips of wires occur.

By the way, I am using an original factory service manual page with a tiny 206 in a circle
on the lower left hand side.
Attachment:
Atwater Kent Shop.jpg
Atwater Kent Shop.jpg [ 123.26 KiB | Viewed 167 times ]

_________________
de
VE7ASO VE7ZSO
Amateur Radio Literacy Club. May we help you read better.
Steve Dow
ve7aso@rac.ca


Top
 Profile  
 
 Post subject: Re: Best way to turn 125 VAC to 110 VDC
PostPosted: Sep Thu 14, 2017 2:38 am 
Member
User avatar

Joined: Jul Mon 26, 2010 8:30 pm
Posts: 19892
Location: Annapolis, MD
pixellany wrote:
Is 90 volts adequate for that set? That could be your distortion issue.....

This got lost....

Also, I'd still like to see a measurement of the actual current used by the set. A test at 90 volts will be better than nothing.

Another thought:
The Power supply can be configured to put filaments and B+ on different filter circuits. Presumably, the filaments can have higher ripple than the B+. If not, then the separate circuits can each be optimized for the load.

_________________
-Mark http://pixellany.com

"It's always something". --Gilda Radner (1946 - 1989)


Top
 Profile  
 
 Post subject: Re: Best way to turn 125 VAC to 110 VDC
PostPosted: Sep Thu 14, 2017 2:59 am 
Moderator
User avatar

Joined: Feb Sun 01, 2009 2:56 pm
Posts: 8844
Location: Victoria, Australia
Just on the pixellany comment: I recently serviced a Roberts R66 mains / battery portable, it had filaments on their own rectifier (metal): It has 3 x 2500uF electrolytics on the filament circuit. I see no reason why a salvaged normal radio transformer could be adapted. But that would require mods to feed the filament/ heaters separately.

Marc


Top
 Profile  
 
 Post subject: Re: Best way to turn 125 VAC to 110 VDC
PostPosted: Sep Thu 14, 2017 4:14 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 11573
Location: San Jose, CA USA
I had looked earlier at whether the filaments could be handled differently than the rest of the circuits. Since they are all indirectly heated cathodes, one might even suspect they could be run on AC. However, in looking over the circuit carefully, it looks like they use the heater string as a convenient voltage divider for all kinds of things, so it not only has to stay DC, but probably even needs to be pretty clean DC. Could still be worth looking into, however, as filtering would be much easier if the filaments were fed off the first filter cap and the rest off the second.

I agree 90 volts might be a little modest for this set. Probably works OK, but could be the cause of distortion.

_________________
Tom K6VL


Top
 Profile  
 
 Post subject: Re: Best way to turn 125 VAC to 110 VDC
PostPosted: Sep Thu 14, 2017 5:34 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 11573
Location: San Jose, CA USA
Here's a quick and dirty bench test of the power supply done with a bridge and capacitors directly off the line.

The circuit is being used to light a 60 watt bulb. The circuit consists of:

- power cord directly into plug (no variac or transformer)
- 3 amp slow blow fuse
- 10 ohm current limiting resistor
- bridge rectifier of four 1N4007 diodes (rated at 1 amp average current)
- two pairs of two 200 uF 160 V electrolytic capacitors (equivalent to two 400 uF caps)
- power resistors between the two caps (two 200 ohms 10 watt and one 300 ohm 5 watt in parallel, for a total of 75 ohms @ 20+ watts )
- voltmeter measuring DC voltage across bulb (and across resistors for current check)

Here it is working, plugged in directly to the AC line:

Attachment:
60 watt bulb at 110 VDC.jpg
60 watt bulb at 110 VDC.jpg [ 115.58 KiB | Viewed 151 times ]


And here's the ripple voltage (using an isolation transformer here, since my scope probe is grounded):

Attachment:
Ripple at 60 watt bulb.jpg
Ripple at 60 watt bulb.jpg [ 82.58 KiB | Viewed 151 times ]


The AC line voltage is 121.5 VAC at the outlet. DC voltage at bulb is 110.5 VDC. DC current, measured with DC voltage across the 75 ohm resistor, is 480 mA (very similar to what the A-K 82-D probably draws in its steady state). Ripple at this current is about 200 mV peak-to-peak, which is probably good enough for the radio.

I also repeated the experiment with a 100 watt bulb and 50 ohms total resistance. This also did not blow either the diodes or the 3A input fuse, and delivered something like 700 mA to the bulb without blowing the diodes.

So what's clear here is that the circuit as proposed is fine for powering a radio like the A-K 82-D in the steady state (after it warms up). While it's warming up, the tube filaments draw a much higher current, probably pushing the total current draw to 2 or 3 amps, which was enough to blow the diodes in the original poster's attempt at this. Larger diodes would solve this problem. Probably just 3 A diodes or so would be big enough.

During warm up, voltage to the radio will be low, current will be high, and ripple will be high, but this doesn't matter at all. Only consequence is that the tubes get a nice soft start, taking a little longer to warm up. Should be just the thing for you guys that like to add inrush current limiters to your radios...

By the way, if for some reason you want to use larger filter caps for even lower ripple, the turn-on current surge won't be any higher. It is limited by the current limiting resistor, not the size of the cap. Size of cap only governs the time length of the surge.

_________________
Tom K6VL


Top
 Profile  
 
 Post subject: Re: Best way to turn 125 VAC to 110 VDC
PostPosted: Sep Thu 14, 2017 5:55 am 
Moderator
User avatar

Joined: Feb Sun 01, 2009 2:56 pm
Posts: 8844
Location: Victoria, Australia
#33 is a filament tube and looks like it relies on DC for bias G3 is on F- (pin 5) and F+ pin1. What has to be watched with the low thermal inertia filament, is any ripple on the filament is liable to e amplified.

Marc


Top
 Profile  
 
 Post subject: Re: Best way to turn 125 VAC to 110 VDC
PostPosted: Sep Thu 14, 2017 6:24 am 
Member

Joined: Sep Thu 23, 2010 6:37 am
Posts: 9084
Location: Powell River BC Canada
The output tube , 2A , is a 33. It is a filament tube, and has a 2nd AF filament shunt
resistor. The tube filament draws 0.26 amps, 2 volts.

The total filament drop is 6.3 x 4 + 2 or 27.2 volts, or ~ 8 watts. That leaves 110 - 27.2 or 88.2 volts
at approx 0.3 amps, or 24.6 watts to be dissipated from the filter choke, RF coils#1 and 2 ,
and filament series resistors , No. 1 and 2 . The pilot light is across No.1. .

Then the thermionic load of the tubes has to be added to the field coil power.

_________________
de
VE7ASO VE7ZSO
Amateur Radio Literacy Club. May we help you read better.
Steve Dow
ve7aso@rac.ca


Top
 Profile  
 
 Post subject: Re: Best way to turn 125 VAC to 110 VDC
PostPosted: Sep Thu 14, 2017 12:07 pm 
Member

Joined: Jun Fri 19, 2009 6:34 pm
Posts: 6970
Location: Long Island
Not impressed. First off, you are underestimating the load taken by the radio, which I know because I owned one. Its around 0.6 amp. They have a couple of resistive voltage dividers in there across the line in addition to heater/filament load, speaker field, and B+ load. The radio actually drew a few milliamps more when it was new because of the wet electrolytic filters, but we can safely assume they're not part of the picture at this point.

Under a load of 0.6 amp, the output voltage of your circuit will fall to about 85 volts and the ripple will increase to about a volt which will definitely be noticed in the speaker. For best results the ripple should be 100 mV or less. Also, the power dissipation in the filter resistors will exceed what you have there. Not sure about the ripple current ratings of your filter caps but I suspect the first cap(s) after the rectifier may get into trouble there too.

If somebody turns the radio off, the output voltage of your circuit rises to 168 volts. Turning the radio back on across the second 200-uF of caps charged won't launch any tubes into orbit but it sure isn't being gentle on them either. The 160-volt ratings of your caps are a bit marginal to deal with a no-load contingency.

The OP already has an excellent answer, which is a switched mode power supply (or actually three of them in series) to get close to the right voltage. If he could come up with one more supply I think he'd be in a very good place.

_________________
"Hell, there are no rules here--we're trying to accomplish something!"

Thomas A. Edison


Last edited by Chris108 on Sep Thu 14, 2017 12:52 pm, edited 1 time in total.

Top
 Profile  
 
 Post subject: Re: Best way to turn 125 VAC to 110 VDC
PostPosted: Sep Thu 14, 2017 12:45 pm 
Moderator
User avatar

Joined: Feb Sun 01, 2009 2:56 pm
Posts: 8844
Location: Victoria, Australia
The over voltage surge due to it being switched off at the set, not the PSU is one I overcame on a PSU with 5V 750mA rail and its B+ @100V and another @ 45V.

The 5V regulators were fitted with 1K resistors, so that they regulated & kept control & the same was done with HV regulators. Shunting say an LR8 with a pass transistor like TIP 50 should solve that surge issue.

Marc


Top
 Profile  
 
 Post subject: Re: Best way to turn 125 VAC to 110 VDC
PostPosted: Sep Thu 14, 2017 2:24 pm 
New Member

Joined: Apr Wed 19, 2017 4:02 pm
Posts: 14
A 100W bulb in series with his circuit will limit the initial surge and keep the things safe.


Top
 Profile  
 
 Post subject: Re: Best way to turn 125 VAC to 110 VDC
PostPosted: Sep Thu 14, 2017 4:25 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 11573
Location: San Jose, CA USA
Chris108 wrote:
...you are underestimating the load taken by the radio, which I know because I owned one. Its around 0.6 amp. They have a couple of resistive voltage dividers in there across the line in addition to heater/filament load, speaker field, and B+ load.

Can you point out the voltage dividers on the schematic? In order to add 100 mA, the total resistance, all the way from the 110 VDC to ground, needs to be 1100 ohms or less. I don't see any resistors like that on the schematic. So I still suspect my current estimate is in the right ballpark.

Chris108 wrote:
Under a load of 0.6 amp, the output voltage of your circuit will fall to about 85 volts and the ripple will increase to about a volt which will definitely be noticed in the speaker. For best results the ripple should be 100 mV or less. Also, the power dissipation in the filter resistors will exceed what you have there. Not sure about the ripple current ratings of your filter caps but I suspect the first cap(s) after the rectifier may get into trouble there too.

As I explained many times, adjust the size of the resistor to get the right voltage. Use a big enough one to handle the power, or use a bucking transformer. If your caps can't handle the ripple current, you bought the wrong caps. If the ripple is too high, use bigger caps or use a regulator. Your claim that this radio needs 100 mV or less ripple seems unlikely, but if it turns out to be true, that's OK.

Chris108 wrote:
If somebody turns the radio off, the output voltage of your circuit rises to 168 volts. Turning the radio back on across the second 200-uF of caps charged won't launch any tubes into orbit but it sure isn't being gentle on them either. The 160-volt ratings of your caps are a bit marginal to deal with a no-load contingency.

Just read what I wrote. Power switch of the radio turns the power supply on and off. Radio load always present. Problem solved.

Chris108 wrote:
The OP already has an excellent answer, which is a switched mode power supply (or actually three of them in series) to get close to the right voltage. If he could come up with one more supply I think he'd be in a very good place.

Lots of ways to build a power supply for this radio. Switching power supplies might make a lot of RF noise, but if they're OK in terms of noise, they're a good solution.

Chris108 wrote:
Not impressed.
Main reason for my posting is to correct egregious mistakes you made in your posts. Should we quote some of them so people can see them again? I know from your many other posts that you're a perfectly capable guy technically. For whatever reason, not all the wheels were turning when you made your posts in this thread.

_________________
Tom K6VL


Top
 Profile  
 
 Post subject: Re: Best way to turn 125 VAC to 110 VDC
PostPosted: Sep Thu 14, 2017 5:33 pm 
Member

Joined: Nov Mon 22, 2010 12:05 am
Posts: 733
Location: Denver, CO
Tom Albrecht, I'm having trouble with your circuit. If the line voltage is 125 VAC RMS, then the half rectified voltage is 62.5 V. That is an RMS value. The peak value is 88 VDC after filtering. How could your circuit generate 11- VDC?
Dave

_________________
I sell old radios at a loss. But I make up for it with quantity.


Top
 Profile  
 
 Post subject: Re: Best way to turn 125 VAC to 110 VDC
PostPosted: Sep Thu 14, 2017 5:37 pm 
Member
User avatar

Joined: Jul Mon 26, 2010 8:30 pm
Posts: 19892
Location: Annapolis, MD
May I offer a bit of simple-minded logic?

The ubiquitous full-wave power supply using a tube and a C-R-C filter has been proven to work. It has also been proven using S/S diodes.

No transformer? compensate with a bit of R on the AC side of the diodes
S/S diodes? ditto...maybe

And, maybe the most relevant to this thread: The design is scalable--assuming we follow engineering 101 and get the right parts.


Suppose I am some random set that wants to be powered from 120VDC. I use 200mA. Now someone wants to install me into a large heater that also uses 200mA. (My power supply is in another box somewhere)
So, the local shop comes an installs a new supply--good for 400mA at 120volts. Yes, things like surge current are higher, but I only see my share. In short, I'll never know the difference.

_________________
-Mark http://pixellany.com

"It's always something". --Gilda Radner (1946 - 1989)


Top
 Profile  
 
 Post subject: Re: Best way to turn 125 VAC to 110 VDC
PostPosted: Sep Thu 14, 2017 5:44 pm 
Member
User avatar

Joined: Jul Mon 26, 2010 8:30 pm
Posts: 19892
Location: Annapolis, MD
dgonshor wrote:
Tom Albrecht, I'm having trouble with your circuit. If the line voltage is 125 VAC RMS, then the half rectified voltage is 62.5 V. That is an RMS value. The peak value is 88 VDC after filtering. How could your circuit generate 11- VDC?
Dave

A rectifier is a peak detector.**

125 RMS is 177peak

the "rectification efficiency" is determined by the circuit impedances, diode drops, etc. and the load current and the size of the first cap.
Suppose you have 70% efficiency---you'll get 124 VDC


To be sure, if you hook up a FW rectifier with no caps and very light resistive load, you'll see the rectified sine wave---but the RMS voltage will be the same as the input.
<<EDIT--fixed a boo-boo>>

_________________
-Mark http://pixellany.com

"It's always something". --Gilda Radner (1946 - 1989)


Top
 Profile  
 
 Post subject: Re: Best way to turn 125 VAC to 110 VDC
PostPosted: Sep Thu 14, 2017 9:44 pm 
Member

Joined: Nov Mon 22, 2010 12:05 am
Posts: 733
Location: Denver, CO
Let me try this one more time. 125 Vrms is 176 volts peak to peak. The bridge rectifier makes each half cycle positive, or 88 volts peak. Understand?
Dave

_________________
I sell old radios at a loss. But I make up for it with quantity.


Top
 Profile  
 
 Post subject: Re: Best way to turn 125 VAC to 110 VDC
PostPosted: Sep Thu 14, 2017 10:03 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 11573
Location: San Jose, CA USA
Hi Dave,

Here's how this works: If the line RMS AC voltage is 125 volts, then the peak voltage is 177 volts. The peak-to-peak voltage is 354 V. Both the positive-going and the negative-going half cycles have an amplitude of 177 V. The bridge rectifier flips all the negative-going half cycles to the positive directions, and lines them all up next to one another. So at the output of the bridge rectifier, you have a train of pulses at 120 Hz, all with 177 V amplitude.

If you now place a capacitor at the output of the bridge rectifier, it charges to 177 V and stays there.

If you then add a load, between the peaks of the cycles, the voltage is pulled downward, and that is the ripple.

A few more details to make everything a little closer to reality: With the current limiting resistor up front, and the 0.7 V drop of each diode, the peaks of the pulses at the output of the bridge rectifier end up being less than 177 V. The big voltage dropping resistor (75 ohms, 20+ watts in my quick and dirty demo) drops another 40 volts. When all is properly taken into account, the output is 110 VDC.

If you decide to build and test this circuit with your radio, you can leave everything as I built it above, with two exceptions:

- use 3 amp diodes or greater in the bridge rectifier (recommend 400 PIV rating or higher)
- use a 75 ohm resistor rated for at least 40 watts
- the amperage of the fuse up front might need to be increased a little (but I think 3 amp slow blow will probably be OK)

You may have to adjust the resistance of the dropping resistor up or down a bit depending on the exact current draw of your radio. I think I'm close with my 500 mA estimate, but it won't be exact.

_________________
Tom K6VL


Top
 Profile  
 
 Post subject: Re: Best way to turn 125 VAC to 110 VDC
PostPosted: Sep Thu 14, 2017 10:34 pm 
Member

Joined: Dec Mon 08, 2008 8:27 pm
Posts: 7926
Location: alameda,CA
So here's a basic question: Was the set having issues operating on 125 AC? I ask because while I realize these sets were all made to operate on 110 AC I've never had a set fail due to being run straight off of 125 AC and that goes not only for many of my older 30's sets but even some of the amps and 50's stereo systems that are used daily. 15 volts isn't really going to make a huge bit of difference when it comes to the set's operation.


Top
 Profile  
 
 Post subject: Re: Best way to turn 125 VAC to 110 VDC
PostPosted: Sep Thu 14, 2017 11:32 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 11573
Location: San Jose, CA USA
It's a DC-only set.

_________________
Tom K6VL


Top
 Profile  
 
Post New Topic Post Reply  [ 78 posts ]  Moderators: Marcc, Norm Leal Go to page Previous  1, 2, 3, 4  Next

All times are UTC [ DST ]


Who is online

Users browsing this forum: Google Adsense [Bot], Yahoo [Bot] and 17 guests



Search for:
Jump to:  




















Privacy Policy :: Powered by phpBB