![]() ![]() |
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
74.190.40.137
Hi
Another primitive question here please. I used 250 ohm/50w resistor to reduce 45v (from V422 TO V377) for my B+ psu. I figured that it will dissipate around 8w. Yet the resistor is so hot i can't touch it. I am using a dale RH power resistor. Is it because I need a heat sink for the resistor? Or is my calculation off?
Thank you.
Follow Ups:
An important fact that nobody has mentioned... the RMS current through this resistor is considerably higher than the total DC current! The current waveform here is not a sine wave - it has a higher peak value due to the charging current for C1. Another to way to look at it - the resistor carries the DC output current PLUS the ripple current in the capacitor.
PSUD will have an RMS value for the resistor in the table - THIS is the figure you must use to calculate dissipation. It's higher than you think...
Hi
Thanks for. your input. This psu is designed after dynamo St 70. C1 47uf I think is not needed at that value. I can lower it to 5 or so to lower the b+???
then I can leave. C2 value as is at 23
I will. Play around.
You can very effectively use PSUD2 to see exactly what sort of voltage difference there will be between 47 uF and 5 uF as the input filter. Go for it Tim.Jeff
Edits: 09/17/12
Timp, can we have a look at a schematic of your power supply? Might there be another way to shed that 45 v without creating a burrito warmer?
![]()
Thanks everyone for helping me. I was up till 1 in the morning to lower the heat.
Here is the PSU design using Duncan's program. The R1 is the burrito heater that's cooking my amp right now.
FIY, this is for my 6bg6g Push Pull power amp using 8k OPT.
Edits: 09/16/12
Did you mean to draw this with a half wave rectifier? If you use a full wave bridge (solid state) and terminate C2 with a 160mA load, the supply will output about 385 volts of nicely filtered B+ for the power tubes. It will unfortunately also exhibit an impedance "bump" at about 15 Hz, but this may or may not be problematic. The bump can be smoothed if you interchange the first two filter caps so that C1 = 23u and C2 = 47u.
--------------------------
Buy Chinese. Bury freedom.
I bet he was "meaning" a full wave center tapped rectifier circuit.
No matter what, I would swap C1 and C2, I agree with you.
I don't think there is any getting around a new power trannie to do the amp properly.
Jeff Medwin
![]()
You guys are amazing! Ok. I come clean! Jeff you are absolutely right. I have a full rect of b+ 386 per side with a center tap. I was lazy that I did not measure the tranny value. From one b+ to the other b+ measures
83 ohm. Did I measure it right?
TRY to "input" into PSUD2 what is really going on, to the best of your knowledge.
Insert a current tap before R2 and assign 150 mA. as its value.
Change the current tap at the end of the filter to 50 mA. OR better yet, add current taps after R2 and R3 to accurately reflect actual current draws, at different locations along the filter.
If 83 Ohms is the DCR across the entire winding, use half of that ( 42 Ohms) as a ROUGH input figure in PSUD, which will be ( not exact but ) an improvement over the 20 Ohms and 83 Ohms you have inputed.
![]()
Hi Jeff
Thanks for helping me. I am extremely thankful. And I learned a lot in the process. I am still confused... I just read voltage on the plate and it says 377.
Somehow, Duncan's PSU says that the reading on C2 is 350ish. So, I may be doing something wrong?
To add more info,
I am using 2 12sn7gt as phase inverter
and 1 12sl7gt as a premp section.
Then 4 6bg6gs as power tubes.
In short, looking at my PSU, do you see any problem using it for a pp amp?
Thanks.
Tim,
That PSUD2 simulation above looks mighty fine to me, and WOW, you have come a long way on this in a very short time.
I would NOT worry about the difference in VDCs of 27 VDC on a 350 VDC supply, that is less than 10%. Is 377 VDC what you wanted ?
What you did is totally opposite of how I would approach the filter to the finals. See my NEW post above, " Pinging ...timp."
Cheers,
Jeff
Question,
Are your Finals (output tubes) fed from C2, and your front end tubes fed from C3 and C4 ?
If so, you need to PSUD input the Finals current draw if ( 180 mA??) right after C2, and PSUD input your Front end draws after C3 ( 15 mA??) and C4 ( 5 mA.??).
Jeff. You are right! C2 is the for the power tubes.
You might consider using a 5R4 to drop more voltage in the rectifier. Also lowering the input capacitor value bfore the 6H choke will make a difference. Try 4.7 to 10 uF with a motor run cap. One other idea is to use a CL-150 inrush limiter on each leg of the transformer, as this will also drop a few volts.
Jeff,
Nice post. I agree with much of what you said. However, the large voltage drop in a lovely sounding ( directly heated !!) 5R4 comes with a price to pay, higher Z.
A single 5R4 is best used in a preamp. Too whimpy for a stereo P-P amp seeking to play back dynamic expressions.
Jeff
The addition of 250 Ohms resistance after the rectifier tube, and the use of a 122 Ohm DCR choke, prior to the Finals, are causing a loss in dynamic capability of your audio amplifier.
There is NO combination of chokes ( that I can see ) which will get you close to the 377 VDC you seem to want to use, when you start with a 386 VAC 20 Ohm power transformer, IF you seek to retain a tube rectifier.
Off the top of my head, to optimize the amp would take a lower VAC secondary on the Power Transformer. Ugh, got milk ??
Jeff Medwin
Right now the 4 6bg6g are drawing 150ma.
Have a look at the datasheet for such resistors. Big difference with a heatsink.
In any case for power resistors I use a part no less than double the wattage.
These are really tough resistors. The chart says an unmounted RH-50 can operate at 20W up to 25° C ambient. I'd still mount it on the chassis like Jeff suggested.
![]()
--------------------------
Buy Chinese. Bury freedom.
Edits: 09/16/12
Thanks TK!
So, 25 C comes out to be 75.2 F. Inside the amp the temp may be much higher. Therefore, mounting is a must! Even then, it runs pretty hot, like untouchable!
On the other hand, a 14 watt resistor dissipating 8.1watts is not going to melt.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
The RH series needs a substantial heatsink. I would also be careful using those for high voltage unless you're sure of the rating. If you'd rather use a standard wirewound power resistor in place of the RH, the link below is a good place to look. Check part no. W250WL8103 (250 ohm, 60W) near the bottom of the page. There are several others on the page that also look promising.
--------------------------
Buy Chinese. Bury freedom.
TK
Thank you for the reply. I have a 13w wirewound in my hand. Do you think it is safe to use that for 8 watt dissipation?
Thanks again for the information.
No-- there are many reasons, but just use a 25 watt-- as a minimum.
As TK indicated, keep voltage across it way below the rating.
---Dennis---
I do think the 13W will be safe, but maybe on the edge for longevity. I try to derate 50%, sometimes more in tube gear. See my other reply below also...
--------------------------
Buy Chinese. Bury freedom.
Sounds like I will be mounting this sucker to the chassis. Or should just add a diode? I got whole bunch of diodes in my hands as well.
Thanks again.
Add a diode? You would need more that one to drop 45 volts.45/.7 = 64 diodes! Sort of impractical.
Edits: 09/26/12
Gees!
You guys are absolutely right! To ensure 50w rating i need to have a heat sink size of 12" × 12" × 0.125" thick aluminum panel (294 sq. in. surface area). I need to add a kitchen sink into my psu!
That's crazy!
Relax,As far as what T_K mentioned on voltage rating, well, you are only dropping 45 VDC across the resistor, so its not a concern.
I would use the 50 Watt resistor bolted to your existing chassis, because you do NOT need to dissipate 50 Watts, only 8.1 Watts continuous.
Also, you can use the chassis to keep your Chipotle Burrito warm. Yumm.
Another alternative, use two 500 Ohm 13 Watt power resistors in parallel, to become 250 Ohms at 26 Watts rated. I sonically prefer Mills MRA-12s, twelve watters, but if in series to the Output tubes, I always prefer near zero or minimum resistances !!.
Jeff Medwin
Edits: 09/15/12 09/15/12 09/15/12
"you are only dropping 45 VDC across the resistor, so its not a concern."
Except that the OP said the resistor has more than 400V on it. Looks like it's safe though; the datasheet says the RH050 has a 2,000 Vrms dielectric strength. Given that it's only dissipating 8W, your idea of bolting it to the chassis seems like a good solution.
--------------------------
Buy Chinese. Bury freedom.
If the Dale is enclosed in an aluminum casing, with two tabs on each end that allow you to mount it, and has some cooling fins, it is ONLY rated 50 Watts when properly heat sinked to a chassis or sufficient material to "sink" the excess heat into. Without the proper mounting, it is no wheres near 50 Watts rated. Most likely it is 20 Watts maximum when not heat sinked.There is a white paste you use, (thermal grease or heatsink compound are common names), applied between the resistor and the chassis contact area, to more easily transfer heat from the device to the heat sink area.
Often, the type resistor I describe is called a "chassis mount" power resistor.
If this resistor is reducing B+ to the output tubes, there are far better ways to accomplish this, but that is another story for another day!
Jeff Medwin
Edits: 09/15/12 09/15/12
Jeff
Thanks for the reply! Yes, it is housed in aluminum enclosure that has two screw holes. I will definitely try to find a heatsink and mount it.
I did not know that 8 watt dissipation would heat up, let's say a 20 watt resistor. Is there a set ration between wattage/dissipation in resistor?
Thanks again.
This is just Ballparking-- but to add clarity-- that resistor that was designed to have a heat sink-- the 50 watter as the example-- I would rate it at 1 1/2 to 2 watts without the heat sink.Chassis make poor heat sinks for a power resistor that is generaing a magnetic field. Why bolt this distortion generator onto/into your chassis?
The easy way is to use air-cooled power resistors that are already rated well above the wattage needed, and mount these floating-- under the chassis on stand-offs-- keeping them away from anything else.
Holes should be drilled into the chassis-- both below and above the resistors to allow air-cooling vertically. Mount the resistor either horizontally or at a slight angle.
--Dennis---
Edits: 09/16/12
Dennis hi,
Would you mind explaining how said resistor, when bolted to a chassis becomes a distortion generator?
If so, what type of distortion/harmonics,etc., would that be?
Cheers,
W
My only experience with these has been with Dale aluminum-housed, heat-sink types.
These generate a lot of heat in free air with very little power thru them. It amazes me how little it takes if there is no heat-sink.
If you're using a chassis as a heat sink, you are heating that part of the chassis more than other areas. This heat-created area is different in density and temp from other areas.
The heat differences aren't conducive to max. performance of the chassis as a neutral device.
Magnetic fields also differ in the heat-sinked and non-heat-sinked areas.
I do not know all of the possibilities for mischief from this-- but common sense tells me some of them. The difference in sound quality between a uniformly made and operated chassis, and one that has different anomalies in either its construction or in its activation induced from various influences on it-- heat and magnetism in the wrong places being only two of them-- the difference in performance of the circuit is obvious when you're comparing amp A- to, say, amp B-- one with a proper-operated chassis and the other with one or more of these problems-- is evident on listening.
Perhaps someone here would care to elaborate on this-- I'm afraid I have only partially answered your question.
Rest assured that this anomaly DOES EXIST if you allow it.
I have used chassis-cooling for FWB's in low-volt power supplies with very few problems-- In these cases, I used Wakefield aluminum coolers bolted onto a chassis bottom that was also aluminum-- the chassis bottom had been cleaned and as much as possible, the aluminum heat sink bonded with it.
As voltages get higher (as in tube plate supplies, etc.), I would tend to avoid letting any part of a circuit get anywhere near a metal chassis.
This attention to detail is far more important in a S.E. amp than it is in a Push-Pull circuit.
---Dennis---
I see the merit of mounting the resistor with good clearance (on standoffs) outside of the chassis based on heat dissipation reasons, but I am at a loss understanding how the unevenly heated chassis contributes to distorsion...
Certainly, any AC circuit is better kept away (cleared, shielded, etc.) from any other circuit, especially signal carrying ones. I see the merit in doing this. Particularly when common mode rejection is small or zero.
I would appreciate if other inmates with similar experiences would pitch in and explain, especially the heated chassis and distorsion part...
Thanks,
Radu.
I'm really confused; the heat is generated based on I^2R. Or, of course if the Vmax of the resistor is exceeded, but that's a different story line.
A low voltage heater or cathode supply (both high current) has the ability to cook the resistor at much higher temperatures than a plate supply.
We're referring to, AL housed resistors and "commonly" affixed to an AL chassis. I'm not seeing the magnetic field?
What part of the circuit is more sensitive, albiet SE vs. PP?
FWIW, I use the RH for heaters sized to NOT cook and bolted to the chassis.
Cathodes get Mills (Huntington) NI 50w bolted to the chassis. When switching from MRA's in series to a single 50w bolt style, I couldn't not hear any difference. Ample convection of air movement around them is something I do address.
If were talking an extreme audible difference, an albino monkey coffin in these positions, will surely make your ears bleed.
I learned so much going through this thread. It is my experience too that RH resistors run pretty hot. I have them in my pre and phono stage. In any case, I am thinking that the resistor is not dissipating more than 11-13 watt. So,just for the sake of it, I bought a 50 watt resistor. I am going to solder it and see how hot it gets.
If you are dissipating 13 watts, wether you use a resistor rated for 25 watts or one rated for 50 watts you are going to create 13 watts of heat. The difference is in the area used to dissipate that heat, correct? In the case of a larger resistor, the heat is spread over a larger area.
I'm guessing that if no air flow is provided, given enough time the 50 watt resistor will get as hot as the 25 watt resistor (asuuming the same power).
I've been amazed at the small size of modern Kiwame 2 watt carbon film resistors, as compared to the carbon resistors of 30 years ago with the same power rating. But these little buggers get HOT!! No touchy.
Same with mills mra. Especially NI resistors due to the construction.
cook an egg.
Without calculating the ambient temp. into the actual heat dissipated, they can and will get very hot. Not had one fail yet though.
FWIW, I kinda got away from the finger tests for heat.
My preferred method is a contact probe.:-)) It doesn't blister.
Love your sense of humor!
S.E. amps have no common-mode rejection, so any distortions introduced into their simple circuit gets amplified.
The reason a good one sounds so good is that the builder has learned to minimize adding distortions into the circuit, and the fact that since S.E. can't reject common-mode, it also can't reject parts of music.
Aluminum on aluminum isn't so bad-- that's OK, the resistance will generate heat and some magnetism when current flows thru the resistive element.
It's better to isolate this as best you can.
---Dennis---
The reason a good one sounds so good is that the builder has learned to minimize adding distortions into the circuit, and the fact that since S.E. can't reject common-mode, it also can't reject parts of music.
....Would the "part of the music it can't reject" stated above be mains ripple or mains x2 ripple depending upon the rectifier used when NOT utilizing critical L?
What has the builder implemented that minimizes the added distortions?
Proper tube topology > finals? Avoiding Miller C, A/C filament supplies away - far and away....
Keeping in line with a good sense of humor, analogously, I sound better in the AM because of 8oz's of Prune Juice at bedtime.
Cheers,
W
"the fact that since S.E. can't reject common-mode, it also can't reject parts of music."
Can you explain what this means? Failure to reject common mode power supply noise is no guarantee of amplifier fidelity.
--------------------------
Buy Chinese. Bury freedom.
Power equals Voltage times Current (in Amperes), Ohm's Law.I always make sure there is 2 to 1, rated versus dissipated, and I am MOST HAPPY when it is 4 to 1, rated versus dissipated, for long term reliability and fail-safe operation.
I ventilate my chassis on DIY builds, above and below power resistor bodies that dissipate a lot. I drill 3/8ths inch vent holes on 9/16ths inch centers, three in a row, for Mills MR-12 power resistors, looks "uptown".
Jeff Medwin
Edits: 09/15/12
...... "Power equals Voltage times Current (in Amperes), Ohm's Law" is not Ohms law. Ohm's law: I=V/R more or less.
it's a combination of Joule's first law and Ohm's law, but all of those relationships between I, V, R, and P are commonly referred to as "based on Ohm's law..."
I ^2 R will give you the actual Pdiss in watts across the resistor.
or
Ohm's law: I=V/R , as Steve O pointed out.
Then x3 or x4 that result, for the resistor rating in watts for good sleep factor and longevity.
NI wirewounds run much hotter than inductive. Mills used to post the curves on their website prior to being bought out by Huntington.
The Dale RH AL. housed resistor(or any AL housed resistor) should have a "W" rating for mounted and not mounted. See how low the wattage rating drops when not properly heat sinked to a metal housing. Some x-fer grease helps.
Cheers,
W
Useful reference, on my wall....
Thanks Jeff!
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: