| porkchop61 |
Hi all-
I was wondering if someone could give this PSU a look over to see if there is anything obviously wrong with it? It looks ok to me, but I'm still learning.
I saw this preamp from Joel on another thread, and it looked simple enought to make as I have most of the parts.
My difference is I'm using a PT from an old Hallicrafters S-40A (don't worry, it's my spare parts donor and in no way pristine).
The spec's on the transformer are 330-0-330 secondary @67mA, so it has a little more voltage than I need. This is the reason that I had to change the original power supply posted with this schematic.
Thanks in advance.
Glenn
Here's the PSU:
Here's the circuit:
 |
|
|
| SY |
| I'm too lazy to calculate the ripple. What did the sim predict? |
|
|
| porkchop61 |
I didn't know that the PSU Designer did ripple calculations :confused:
I was more concerned that the original PS had the 20 & 100uF capacitors reversed. I thought that 100uF was too large for a 5Y3GT rectifier.
Glenn |
|
|
| SY |
Well, you have 1.5K between the rectifier and the input cap, so you're unlikely to have an issue- the ripple current is limited by that resistor.
To get ripple, look at the voltage across C2/I1 after the initial settling in (a 5 second reporting delay will do fine). The sim gives min, max, and the difference. That difference is the ripple. |
|
|
| porkchop61 |
Hi Sy-
Min is -52.962mV, max is 273.65 Delta is 273.7
Does that mean the ripple is 273.7V???? Something doesn't seem right.
If I zoom in on the graph of the voltage at I1, I only see ~0.1V ripple.
Glenn |
|
|
| SY |
| No, it means that you didn't put in a reporting delay time. Use 5 seconds, then you'll get sensible ripple numbers. |
|
|
| porkchop61 |
OOOPS! Sorry, you did say that :o
Yes, that's better, now I get 84.947mV
Smooth enough for a preamp?
Glenn
BTW: V. Sattui & Domaine Chandon are my 2 most favorite places to vist when I'm there :D |
|
|
| porkchop61 |
Interesting, when I swap the 2 capacitors (100uF then 22uF) I get 46mV ripple.
This is the way it was in Joel's original schematic. I guess it was right after all.
Glenn |
|
|
| SY |
Well, it's all a matter of the S/N ratio you want to achieve. What you need to do is calculate the expected noise from the power supply rejection. In the case of a cathode follower, the rejection is approximately mu, so with a 6SN7 (mu = 20), you'd expect the ripple to contribute about 2.3mV of noise.
I don't know the 12AY7 specs offhand (I'm not at my own computer at the moment so I can't grab the data sheet), but what you want to do is get the plate resistance and mu at the tube's operating point. With those numbers in hand, you can see that there's a voltage divider formed by the load resistance (essentially, the 100K plate resistor) and the tube's source impedance, rp + (u+1)Rk, where Rk is the cathode resistor. Using a tube whose specs I do know from memory as an example (12AX7), the plate resistance is about 60K, the mu is about 100, so the source resistance would be 60K + (101) 1.5K = 210K. In that case, the rejection would be something like 210/(100 + 210), i.e., not very good. So you'd get a contribution of another 30mV or so from that stage. Referenced to a 3Vpeak output, that's a S/N of about 40dB.
Hopefully, the 12AY7 will calculate out better.... |
|
|
| porkchop61 |
Thank you very much for that information.
I'll have to look at the spec's and do the math.
I thought this was the case, how much you could live with at the speaker end, after all, it is an amplifier :)
Thank you.
Glenn |
|
|
| SY |
I'm of the opinion that S/N ought to be better than -100dB. After all, we have sources with that kind of dynamic range, so why let the line amp throw away what our front-end eforts have given us?
In the case of my preamp (the Heretical), the ripple-induced noise is down in the microvolt range. Subjectively, I feel that I can pick up more low-level detail because of that. Fortunately, it's not hard to get ridiculously low levels of ripple. |
|
|
| Joel |
Oh god, what is the POINT of this amp?? It's got NFB, too much gain, capacitor coupling, and parallel tubes...
I guess three years can make a big difference. :( |
|
|
| porkchop61 |
You started it Joel ;)
If this amp is not an "acceptable" design, any other suggestions using 6SN7 tubes? I need more than unity gain though as I 'd like to use this amp for headphones as well as a preamp.
Glenn |
|
|
| burnedfingers |
Having built this a few years back as my first project I can now look back and say it was ok but nothing to write home about.
I was also experiencing power supply ripple at the time and questioning the design. The answers I was getting regarding the hum I was experiencing was there was a grounding problem (mine). The only thing I can say in its favor is it is different. Any 12B4, 6SN7, 5687, or probably anything else will kick its butt to the curb. My suggestion is to move along to the "Aikido" and build it. It is great and holds the promise of tube rolling with great results. There is a power supply that has low ripple that is well thought out. |
|
|
| burnedfingers |
| quote: | Originally posted by porkchop61
You started it Joel ;)
If this amp is not an "acceptable" design, any other suggestions using 6SN7 tubes? I need more than unity gain though as I 'd like to use this amp for headphones as well as a preamp.
Glenn |
Check thru the past threads and you will find a 5692 line stage that Frank posted. I built it and its very nice. |
|
|
| burnedfingers |
Maybe this will work.
I built the circuit on top. |
|
|
| porkchop61 |
Thank you!
That looks simple also.
I'll look at the Aikido, but my goal was to build something with the parts on hand (6SN7's, 330V PT, etc.).
Joel was kind enough to send me this suggestion also.
Glenn |
|
|
| porkchop61 |
Thank you.
The 12SN7 is the same as the 6SN7 except for the heater voltage, correct?
The tube count is going up then for a stereo version of this circuit (4 tubes).
Glenn |
|
|
| Joel |
Although I didn't design that PSU. But I do think the direct coupled circuit would be an improvement. It's also simpler.
My gripe about both circuits - and any other RC coupled design used to drive a low impedance like headphones - is that they rely on the load to generate the NFB which improves the linearity. With 100 ohm phones, you essentially have a 6SN7 (or two) with a 100 ohm load. What do you think that's going to sound like?
The fact that a circuit like this sounds almost identical to the headphone amp on the player itself says a lot about the equally poor quality of the solid state version, and even more about the lack of power needed for most 'phones.
There really is only one way to drive a low impedance (probably reactive) load with vacuum tubes, and do it linearly, and that's with a transformer.
Save the cathode follower for buffering a high source resistance.
Joel |
|
|
| porkchop61 |
Please refer to post #17 for the schematic.
Can I use an electrolytic capacitor for the 1uF capacitor on the output of this preamp? I know, someone will say "yuck" for using an electrolytic in this position, but I only ask because I want to bread board this circuit, and all I have are electrolytics.
Is polarity important if I decide to use them?
I will use a "Kimber Kap" if this works ok.
Thank you.
Glenn |
|
|
| martinab2 |
Lots of interesing linestage ideas here
I built the 7193 one about half way down the page - very simple but sounds good. Might be of interest to you as a 7193 is more or less half a 6SN7. |
|
|
| porkchop61 |
Thank you for that interesting reading.
I really like the tubes with the 2 grid caps! Very unique.
I'm glad to hear you like the oil filled motor run caps as I'm building a KT88 SET power amp using those. I'm not finished yet, so I'll have to let you know how it sounds later.
I still need an answer to the electro cap useage for the output line. I've never seen them used in this position, but I don't think it should be a problem. I also don't think the polarity matters, but I'm still relatively new at this.
Thanks.
Glenn |
|
|
| SY |
| Yes, you can use an electrolytic and yes the polarity will matter. Positive end toward the cathode. |
|
|
| porkchop61 |
Thanks Sy!
I see now that the schematic does show the + towards the cathode. It will probably sound like CR_P, but I do have them on hand. It's just for a bench top test anyway.
Glenn |
|
|
| porkchop61 |
I found this circuit which looks very similar to Joel's design (see post #17). Does this design have any merits above Joel's circuit?
Thanks
Glenn |
|
|
| SY |
| Yes, the cathode follower has a much more sensible load and biasing. However, you'll want to put a protection diode between the grid and cathode to allow the tube to survive turn-on and turn-ff. |
|
|
| porkchop61 |
| quote: | Originally posted by SY
Yes, the cathode follower has a much more sensible load and biasing. However, you'll want to put a protection diode between the grid and cathode to allow the tube to survive turn-on and turn-ff. |
What is the issue with turn-on/turn-off? Is it a cold heater issue? I'm using a 5Y3GT rectifier with this circuit.
Diode between the screen and cathode of which half of the tube (pins#4&6)? Which direction is the diode installed? Cathode of the diode to cathode of the tube, or the other way around? 1n4007's (solid state!!) ok?
Maybe you could explain why a diode is needed for turn-on/off protection, I'm still learning :)
Thanks Sy! |
|
|
| Joel |
Hi Sy,
Why do you think 18k is a better load for a 6SN7 than 47k? |
|
|
| porkchop61 |
Here's another couple of questions on this circuit. What is the 1Meg resistor tied to ground right before the output for?
What happens if the output capacitor fails and you have your headphones plugged into the amp? :eek:
Glenn |
|
|
| dhaen |
| quote: | Originally posted by porkchop61
Here's another couple of questions on this circuit. What is the 1Meg resistor tied to ground right before the output for?
What happens if the output capacitor fails and you have your headphones plugged into the amp? :eek:
Glenn |
1. To stop a lout pop when the headphones are connected, or a shock if you touch.
2. It won't if you adequately rate it. It's not life threatening so adequate rating is sufficient. |
|
|
| SY |
| quote: | Originally posted by Joel
Hi Sy,
Why do you think 18k is a better load for a 6SN7 than 47k? |
Current. 6SN7 likes current. The 5.6mA number on the schematic doesn't make sense to me; looking at the tube curves, it looks more like 2 or 3 mA.
Porkchop, the diode is connected between cathode and grid, reverse-biased in normal operation (i.e., arrow pointing to the grid). Remember, when the first section is drawing no current (e.g., during warmup), the grid will see the full B+ on it while the cathode is still at ground. The diode clamps the grid and cathode to prevent such an overvoltage situation. |
|
|
| Joel |
"Current. 6SN7 likes current. The 5.6mA number on the schematic doesn't make sense to me; looking at the tube curves, it looks more like 2 or 3 mA."
2.8mA per triode, 5.6mA total. As I said before, Porkchop61 has modified a drawing of mine of the single 6SN7 stage into an amp of his own design. The labels got moved around in the process, a second 6SN7 channel got added, and a PSU.
RE: 6SN7 and current
Linearity is affected by op point and load. 18k is a poor load for a 6SN7, regardless of the Ib. If you want to run the standard 9mA, 250V point, then switching to a choke cathode load like the Lundahl LL1667/15mA is a fine choice.
But even with a choke, or the world's greatest CCS, at the end of the day you're still going to be loading the 6SN7 with whatever you hook up to this thing, and not truly the choke/CCS/resistor.
Joel |
|
|
| SY |
Well, that's point- the AC load will be something more like 10K. The current swing to get 2.5VRMS across 10K is 0.25mA (or more than a third of a milliamp peak). That certainly suggests that a high idle current would be more desirable so that the signal current is a much smaller percentage of the idle current.
The rp and mu level out quite a bit when you run higher currents (more than 5-7mA). |
|
|
| porkchop61 |
| quote: | Originally posted by dhaen
1. To stop a lout pop when the headphones are connected, or a shock if you touch.
2. It won't if you adequately rate it. It's not life threatening so adequate rating is sufficient. |
Yes, I thought this might be the case. I use this resistor on all my guitar pedal builds as a pulldown resistor for the same reason.
Ok, good to know I won't get toasted if the capacitor fails :eek:
| quote: | Originally posted by SY
Porkchop, the diode is connected between cathode and grid, reverse-biased in normal operation (i.e., arrow pointing to the grid). Remember, when the first section is drawing no current (e.g., during warmup), the grid will see the full B+ on it while the cathode is still at ground. The diode clamps the grid and cathode to prevent such an overvoltage situation. |
Thank for that explanation Sy, makes sense to me.
Time to add those 1N4007's to the schematic :)
Glenn |
|
|
| porkchop61 |
Sy-
With regards to the diode you suggested.
What did they do 50 years ago? They didn't have soild state diodes :scratch1:
Just curious.
Glenn |
|
|
| SY |
| They often used neon bulbs. You'll see them a lot in old tube opamps and in amps with direct coupling like the Marantz 9. |
|
|
| porkchop61 |
Ahh, yes I have seen neon lamps in old guitar amplifier circuits.
Very creative use of a neon lamp.
Thanks.
Glenn |
|
|
| porkchop61 |
Is this correct for the diode installation?
Glenn |
|
|
| Joel |
No, the 1N4007 should be reversed. The cathode of the diode should be to the cathode of the tube. When the tube warms up, the voltage at the tube's cathode will be greater than at its grid, reverse biasing the diode and removing it from the circuit.
Also, if you plan on using this circuit to run typical 300 ohm headphones, you will need an output capacitor at least 100uF. For 150 ohms, at least 220uF, etc.
Joel |
|
|
| porkchop61 |
Ok, try this on for size.
Joel-
What if I want to run both headphones and an input to a power amplifier for this preamp? Can I split the signal path after the cathode, and go to 2 different capacitors? One for an RCA output for a power amplifier, and one for a headphone jack? 100uF-220uF we're now talking electrolytic caps only right? I don't think I've ever seen a 100uF 250V poly cap.
Thanks for your help guys.
Glenn |
|
|
| Joel |
Yes, that's exactly what you would do. Use a SPDT switch, with one leg going to a 220uF electrolytic, and the other to a film cap of 0.22uF or larger.
Joel |
|
|
| porkchop61 |
| quote: | Originally posted by Joel
Yes, that's exactly what you would do. Use a SPDT switch, with one leg going to a 220uF electrolytic, and the other to a film cap of 0.22uF or larger.
Joel |
Hi Joel-
Do I have to use a switch? Is there something wrong with just leaving one circuit open when not in use?
Ok, here's a basic question about the value of the output capacitor. What does the value of this capacitor do to the output of this preamp? Does it control impedance, or is it frequency related. I'd like to understand what effect the value has on the circuit.
Thanks
Glenn |
|
|
| Joel |
If you don't cut off the other capacitor not being used, it hangs off the output directly connected to the cathode. That makes a great antenna.
The output capacitor, besides blocking DC, forms a filter with the resistance of the load in parallel with the 1M resistor to ground.
The -3dB point of this filter is 1/(2Pi*R*C)
Joel |
|
|
| porkchop61 |
| quote: | Originally posted by Joel
Also, if you plan on using this circuit to run typical 300 ohm headphones, you will need an output capacitor at least 100uF. For 150 ohms, at least 220uF, etc.
Joel |
Good morning all (it's 8:40AM here)-
I measured my headphones (vintage 1983 Yamaha studio), and they are ~130 per side. I assume the 220uF elect cap would be ok for these, yes? How do you calculate the output capacitor value for a given headphone load?
Thanks.
Glenn |
|
|
| porkchop61 |
| quote: | Originally posted by Joel
The output capacitor, besides blocking DC, forms a filter with the resistance of the load in parallel with the 1M resistor to ground.
The -3dB point of this filter is 1/(2Pi*R*C)
Joel |
This is another question I had about this circuit. I added the 1M pulldown resistor to the output, originally this circuit did not have this. Can I still prototype the circuit with the 1uF capacitor called for, or do I need to change this value initially now. I will be trying this circuit out by connecting it to a solid state power amplifier initially (vintage Marantz). Or maybe I shouldn't even use the 1M resistor at all since I don't plan on unplugging anything while the unit is powered on.
Thanks
Glenn |
|
|
| SY |
Use the 1M resistor. For a nickel, you can forever not worry about accidental bangs and pops if an output connection accidently goes intermittent or pulls out.
For a solid state amp, the load impedence is generally 10-20K. I'll leave the calculation of the capacitor to you. |
|
|
| ElectricHead |
| What about the phase shift? All circuits shown here converting the phase by 180°. |
|
|
| SY |
It's not phase shift, it's polarity reversal. If one worries about it, swapping speaker wires (or doing any other polarity reversal in the signal chain) cancels it out perfectly.
Worry more about absolute polarity in your source material. That's something that you won't know and can't control, other than by listening and guessing. |
|
|
| ElectricHead |
| Sorry, I meant polarity ofcourse. I think you can hear the polarity reversal in some systems. |
|
|
| porkchop61 |
Okay SY-
Here's my attempt at calculating output impedance for this amplifier. I looked at Mr Aiken's site (he's a great resource)http://www.aikenamps.com/Equations.htm
I just can't seem to match his circuit calculations to mine. My circuit (Joel's) is on the right. I guess it's because the circuit I'm using is different from his example that I'm having trouble.
For instance, he uses the resistance RI in parallel with the plate resistance even though the output is taken off the cathode, is this right? I'm also not sure if I use the 1M resistor as the input resistance of the next stage, or the resistance of the headphones (130 ohm), or the two added together? |
|
|
| SY |
I'm not understanding the question very well- sorry, not much sleep the past few days. In any case, the output Z of a cathode follower will be very close to 1/gm. For a 6SN7 running at 5-6mA, the gm is roughly 2.5mA/V, so the output impedance will be roughly 400 ohms. Is that what you're asking?
For 130 ohm headphone drivers, you might want a lower source impedance than that, or consider using a step-down output transformer. |
|
|
| porkchop61 |
Thanks for looking at that post Sy.
Can you briefly explain "lower source impedance"?
Is this the impedance of the circuit, or the device connected to the circuit ie. a CD player, tape deck, etc.?
Also, this circuit does not have a capacitor on the input, is that a big deal? I guess it doesn't matter as long as there is no DC on the input. I've also seen some of these preamps with a 1M resistor on the input as well.
Sorry for all the questions, and thanks for your time.
Glenn |
|
|
| SY |
Let's start easy with a quick homework assignment. Are you familiar with the concepts of Thevenin and Norton eqivalents? If not, look 'em up in any electronics text, try to work a couple of examples, and we'll discuss.
If you are familiar with them, the source impedance is the Thevenin equivalent resistance of the cathode follower's output, i.e., from the point of view of the load, the cathode follower looks like a Thevenin source with a resistance of 400 ohms. |
|
|
| porkchop61 |
| quote: | Originally posted by SY
Let's start easy with a quick homework assignment. Are you familiar with the concepts of Thevenin and Norton eqivalents? If not, look 'em up in any electronics text, try to work a couple of examples, and we'll discuss.
If you are familiar with them, the source impedance is the Thevenin equivalent resistance of the cathode follower's output, i.e., from the point of view of the load, the cathode follower looks like a Thevenin source with a resistance of 400 ohms. |
Fair enough.
I knew I should have done the double major (ME/EE) in College :)
I'm just a dumb ME, but I hope to remedy that soon!
Thanks.
Glenn |
|
|
| SY |
| Don't worry, this stuff is waaaay easier than mechanics. |
|
|
| porkchop61 |
It's funny, but I can usually equate an electrical component to a mechanical one (It helps me think better that way).
For instance, a power supply capacitor is like a flywheel in an engine. It helps smooth out the revolutions of the crank every time the engine fires. I guess Physics is applicable across many fields of study. That's all it comes down to in the end anyway.
Oh, I forgot to ask, about the large electrolytic capacitors people use for the output of these headphone amps. Are these the snap-mount type ones I see on tops of some chassis? I'm looking for 220uF/450V units, and I don't see them in the "usual" radial or axial packaging. If they are, how do you mount these things? With a clamp like the very large ones for power supply filters?
Thanks.
Glenn |
|
|
| Joel |
| a 250V rated cap on the output should do it for this circuit. |
|
|
| SY |
| Porkchop, slightly OT, but... I'd highly recommend getting a copy of "Similarities in Physics," by Shive and Weber. Despite the rather dry title, it's very readable and enjoyable. The focus of that book is showing that the same equations pop up in differing contexts (e.g., mechanics and electronics), and that there's a sound basis in drawing these analogies. |
|
|
| porkchop61 |
Thanks SY, I'll check it out.
Glenn |
|
|
| porkchop61 |
I almost finished this preamp last weekend, and the B+ was within 2 volts of the predicted value! That PSU Designer II software is awesome!
I thought I'd ask this question even though it probably doesn't mean anything. Is running the 6SN7GT's at 254V ok even though the schematic says 250V? This could be due to high power line voltage right? Regardless, I don't think this is a big deal right?
Joel?
Glenn |
|
|
| Joel |
10% either way on ANY voltage in any tube amp is acceptable. And I'm told that in the old, pre-DMM days, a 20% margin was the limit.
Some old power amp manuals don't even give readings for the small signal tube voltages because they would've had hundreds of guys calling in wondering if their unit was broken because it was a few volts off here and there.
Joel |
|
|
| porkchop61 |
Thanks Joel, I thought as much.
I'll have to post pictures when I get it finished.
Thanks for all your help!
Glenn |
|
|
| SY |
Nice job! BTW, a messy workbnch is one on which actula work is done.
The rtr looks familiar, but I can't place it. Akai? |
|
|
| porkchop61 |
| quote: | Originally posted by SY
Nice job! BTW, a messy workbnch is one on which actula work is done. |
Yes, you can't clean and build at the same time.
| quote: |
The rtr looks familiar, but I can't place it. Akai? |
No, Its' a cheap Sony but it works pretty well. I have two as they were both saved from the dumpster in '78.
I'm not too crazy about the sound of the Chinese 6SN7GT tubes, so I'll probably order a few different brands. Any recommendations on favorite brands of this tube?
Thanks
Glenn |
|
|
| Joel |
Looks good Glenn!
I'm partial to no-nonsense aluminum boxes for my projects too.
As I said before, I think your amp will make a great preamp, but a lousey headphone amp. Tell us what you hear.
I have a Teac 3430 reel-to-reel. It's a pain to setup, but I do like the sound.
Joel |
|
|
| porkchop61 |
Yes, I'm going to have to try it with my Yamaha M-50 power amp for the meantime. I'm still not finished with my KT88 SET power amplifier. Those SE OT's are expensive, and I need two!
Thanks for your help.
Glenn |
|
|
| SY |
| I'm probably the wrong person to ask about currently-available 6SN7s- I've been living off my stock of 5692s purchased long before these were considered audio jewelry. Morgan Jones recommends Pinnacle 6J5GT (single triode version of 6SN7), and seems quite keen on 6SN7 variants with carbonized envelopes. |
|
|
| porkchop61 |
| quote: | Originally posted by Joel
My gripe about both circuits - and any other RC coupled design used to drive a low impedance like headphones - is that they rely on the load to generate the NFB which improves the linearity. With 100 ohm phones, you essentially have a 6SN7 (or two) with a 100 ohm load. What do you think that's going to sound like?
There really is only one way to drive a low impedance (probably reactive) load with vacuum tubes, and do it linearly, and that's with a transformer.
Joel |
Okay Joel-
Now I think I'd like to try a bench test with a transformer for driving headphones now that I have the preamp working. Does a 1.5k primary with around 100 ohms secondary sound useable? Or do I have to match the secondary exactly with my 130 ohm headphones?
I assume this is not a large transformer. Any ideas where to buy one of these? Do I look for "impedance matching" transformer types, or "interstage"?
Thanks
Glenn |
|
|
| SY |
| Look for an impedance matching type with roughly 10K:150 ohm (don't worry about getting things exact). They're not too big, but a good one can be pricey. Check Cinemag to see what they might have- their pricing tends to be pretty reasonable. |
|
|
| SY |
| A mike input transformer will saturate in this application. Something more like a line driver would be suitable. I'll take a quick run through Cinemag's site and see if there's anything suitable. |
|
|
| porkchop61 |
Look at post #75 as I just edited it.
Glenn |
|
|
| SY |
| That looks pretty good. The Cinemag CMO-15/100CT might also be suitable. Neither of these have gapped cores, so you'll have to capacitively couple to them. |
|
|
| porkchop61 |
| I have a 1uF cap on the output now, is that sufficient? |
|
|
| Joel |
Sorry to be a contrarian, but I don't like the idea of using a typical "matching transformer", or using capacitor coupling to it (parafeed).
With an 8k impedance driver like a plate-loaded 6SN7, I would simply use an off-the-shelf 6k-10k primary output transformer, and hook the headphones up to the 16 ohm tap. You could do push-pull, but an SE unit would also be fine at the microwatt levels needed.
There was an excellent thread here discussing just such a design, with good informative posts by PRR. Search for it.
Joel |
|
|
| porkchop61 |
| quote: | Originally posted by Joel
Sorry to be a contrarian, but I don't like the idea of using a typical "matching transformer", or using capacitor coupling to it (parafeed).
With an 8k impedance driver like a plate-loaded 6SN7, I would simply use an off-the-shelf 6k-10k primary output transformer, and hook the headphones up to the 16 ohm tap. You could do push-pull, but an SE unit would also be fine at the microwatt levels needed.
There was an excellent thread here discussing just such a design, with good informative posts by PRR. Search for it.
Joel |
Thanks Joel.
I didn't like the idea of being tied down to one output impedance. What if I want to use my tiny Sony headphones that measure 4 ohms? I assume that the solution you suggested would work over a range of different headphone impedances?
I'll look at that thread you suggested.
Glenn |
|
|
| porkchop61 |
I'd now like to try this preamp with my turntable. It's a Denon DP-60L with a moving magnet cartridge (I lost track of my Signet moving coil cartridge after I broke the needle, doh!).
Will this preamp have enough gain to use with this, or do I need to add another stage? If I have to add another stage, can I add some switching to remove the added stage? I guess It would be easier to just add the input for the turntable before the new stage (if needed), and leave the existing inputs where they are.
Any desing ideas?
Thanks.
Glenn |
|
|
| Joel |
Other than rolling off the bass at 3.4Hz, what's the .047uF cap for?
Also, the 1M resistor that will alledgedly save you from catastrophe should the wiper go bad on the pot - if the wiper goes bad, you'll get no music... which would be a pretty good indication that something is wrong. Does anybody think they'd be running their amp for minutes or hours with an open pot?
In any event, a 6SN7, even with zero bias, is simply not going to destroy itself with 48.5k of series resistance. The extra resistor is silly, IMO.
Joel |
|
|
| SY |
You'll need a separate stage for phono to provide extra gain and RIAA equalization.
For the extra 2 cents that a resistor costs, use the 1M safety resistor. |
|
|
| porkchop61 |
| quote: | Originally posted by Joel
Other than rolling off the bass at 3.4Hz, what's the .047uF cap for?
Also, the 1M resistor that will alledgedly save you from catastrophe should the wiper go bad on the pot - if the wiper goes bad, you'll get no music... which would be a pretty good indication that something is wrong. Does anybody think they'd be running their amp for minutes or hours with an open pot?
In any event, a 6SN7, even with zero bias, is simply not going to destroy itself with 48.5k of series resistance. The extra resistor is silly, IMO.
Joel |
I was trying out the variation on this circuit that BernardB posted here (bottom circuit):
http://www.diyaudio.com/forums/show...&threadid=67433
I thought the .047uF input cap was to prevent any DC from the input (not that you should have any)? Other than that, I don't think having it roll the bass off at 3.4Hz is a bad thing. Kind of like a sub-sonic filter? In that case, why don't designers use something that rolls off the bass at around 10-18Hz? Most speakers can't reproduce that anyway, just curious.
I have also not changed the output cap to the suggested 5uF from the 1uF that I already have in there. Just waiting for them from Angela.
Glenn |
|
|
| porkchop61 |
| quote: | Originally posted by SY
You'll need a separate stage for phono to provide extra gain and RIAA equalization.
For the extra 2 cents that a resistor costs, use the 1M safety resistor. |
Different Avitar SY??
Thanks for that information. Is it possible to incorporate an extra stage into this design, or am I looking at a new build? Any suggestions as to a design for another stage? Maybe a 12AX7?
Can I switch this new stage in or out of the circuit, or like I said just add another pair of RCA jacks for a phono input to this stage?
Glenn |
|
|
| SY |
I like to change my avatars every so often. This one is in honor of my kid's birthday.
Figure that you'll need a whole new build, unless you've got room for 3-4 more tubes. There are a lot of sources for good RIAA circuits- certainly, you can't go wrong by starting with one of the ones from Morgan Jones's book. |
|
|
| porkchop61 |
Thanks, I'll look at that book.
You're son is a future CIA Greystone candidate? (My other passion is cooking).
Glenn |
|
|
| SY |
Unlike most 5 year olds, Jimmy won't recognize most cartoon/merchandising characters, but he gleefully points out pictures of Wolfgang Puck. And at 2, he could name all the Iron Chefs, say what color they wore, what cuisine they represented, and what prop they would hold in their hand at the show intro. He's old enough to do plate arrangements now when I'm putting a nice dinner together. His musical tastes run to They Might Be Giants.
Greystone's not good enough for him; he's going to study at Bocuse if I have anything to say about it! |
|
|
| porkchop61 |
Apologies in advance to everyone for being OT :(
That's Paul Bocuse in Lyon, right?
Very nice, I bet tuition is comparable to reputation :)
I almost went to NECI since I'm in Connecticut, but my Father had other ideas since he was the Dean of the school of Engineering at UCONN.
Good luck, at least you'll have someone to cook nice things for you when you retire.
Back to tubes, I found this phono preamp schematic, and it looks simple.
Glenn |
|
|
| SY |
Simple, yes, but high output impedance and high distortion. I would guess that the noise isn't so great, either. Rule of thumb- if there aren't at least three tube sections in an RIAA amplifier, there's a severe compromise somewhere.
Bocuse's school is just north of Lyon (a suburb called Ecully). I used to date their International Studies director, so I have some familiarity with them. I'm completely self-taught as a cook, but have managed to eke out one or two edible dishes in my day. Jimmy's got some work ahead of him if he wants to impress the old man. |
|
|
| porkchop61 |
| Okay, here's another one I was considering: |
|
|
| SY |
| Not enough tubes. Take a look at Jones's ECC88-based phono preamp. |
|
|
| porkchop61 |
Hi all-
I'm back to the 6SN7 preamp/headphone amp.
Please refer to post #82 for the schematic.
I installed a DPDT switch for the outputs to switch between the RCA jacks and the headphone jack. The RCA uses a 1uf electro cap, and the headphone jack uses a 220uf electro.
The wierd thing is, I'm getting ~130VDC on the output of the 220uf capacitors for the headphones! When I switch to the 1uf caps for the RCA line out jacks, I get 0VDC at the jack.
Any ideas as to what is going on?
Why would the 1uf block the DC, and the 220uf not?
I don't think I want to plug my headphones into 130V!
Glenn |
|
|
| SY |
| Well, you do have a 22 second (probably more) time constant there. It might take a while to settle to zero. |
|
|
| Brian Beck |
| quote: | Originally posted by SY
Well, you do have a 22 second (probably more) time constant there. It might take a while to settle to zero. |
Yeah, and if you plug in your headphones during that time, pop goes the weasel. There is enough stored charge in 220uF to zap a little voice coil. This kind of circuit is risky for sensitive headphones. You could put a shorting relay with a timer on the output to provide a time delay for the big output cap to charge. Or, put in maybe a 2K shunt ground reference resistor instead of the 100K to speed up the charging time. The 2K in parallel with your 32 to 300 ohm headphones won’t significantly reduce levels or increase distortion. Then, and this is a bit brutal, you can add a pair of series zeners (back-to-back) across the output line to clamp start-up transients. For Sennheisers maybe a pair of 8.2 volts zeners; for Grados, lower.
| quote: | Originally posted by Joel
Also, the 1M resistor that will alledgedly save you from catastrophe should the wiper go bad on the pot - if the wiper goes bad, you'll get no music... which would be a pretty good indication that something is wrong. Does anybody think they'd be running their amp for minutes or hours with an open pot? |
That's not the point. It's not the presence or absence of music that we’re worried about; it's the sudden large step-function jump in the DC levels as the grid bias goes from a ground reference via the pot suddenly to an open condition. When open, a grid will find a voltage point of equilibrium where positive and negative grid currents cancel. This is typically between -1.0 and 0 volts of equivalent grid bias. If your grid bias is normally 3 or 4 volts (a wild guess - I didn’t do the load lines on the schematic in post #82), you just snapped a 3 volt transient into your input. The output, of course, will react to that with gain. With headphones on, it might sound like a lightning bolt in your ear. Even worse, if the pot is intermittent, as you turn it you’ll get a series of makes and breaks, and you’ll experience multiple lightning bolts per second. More than one tweeter has been lost through the decades due to intermittent volume pots.
I agree with SY. For a couple of cents, keep the safety resistor even if you drop the input coupling cap. |
|
|
| porkchop61 |
Thanks for the answers.
Yes, I did notice that the DMM I had on the circuit was headed down slowly. I guess I'll go back and give it time to zero-out.
I'm probably going to use a transformer as previously discussed in this thread, but I'm experimenting for now.
That's the fun of DYI and learning for a newbie like me :)
Now if I can decide on OT's for my SE KT88 project I can finish that also! I'm undecided as to who's transformer to use, the "new" Hammond 1628SEA, or the Edcor.
Thanks again.
Glenn |
|
|
| Drewaudio |
| quote: | Originally posted by SY
Not enough tubes. Take a look at Jones's ECC88-based phono preamp. |
Hi, any chance of a url for this phono?
Cheers
Andrew:) |
|
|
| Drewaudio |
Thank you! I see our local library has two copies so I will take a look today.
Kind regards
Andrew:D |
|
|
|
