Most triodes start conducting appreciable grid current before Vgk reaches 0V.
The JFET input opamp I'm using with 25pA (typical) bias current at 25degC is more than good enough for my application (and is pretty much nulled out by the identical resistor on the non-inverting input), and would be more than good enough for measuring the input current of SYN08's bipolar-input phono amplifier.
It's a 10 minute dead-bug job. He could use a junk-box TL071 or similar here.
The JFET input opamp I'm using with 25pA (typical) bias current at 25degC is more than good enough for my application (and is pretty much nulled out by the identical resistor on the non-inverting input), and would be more than good enough for measuring the input current of SYN08's bipolar-input phono amplifier.
It's a 10 minute dead-bug job. He could use a junk-box TL071 or similar here.
syn08 said:
I'm not sure about that. What is the DC resistance of an MC cartridge? I googled and got 5-10 ohms. That will deliver you an input offset voltage of 5-10uV for 1uA of input bias current.
This is still a substantial offset volatge and could be sensed with a chopper opamp based servo, using a device such as the Maxim-ic device 1audio mentioned, which has an input bias current of only 1pA. The typical input offset voltage is a low 0.1uV with only 10nV/degC drift.
Some commercial BJT MC RIAA amplifiers - even very very expensive ones - have a cap in series with the cartridge. I would hesitate to connect a BJT MC RIAA amplifier to a say 1000 USD cartridge...
If a few R-core transformers can fix the current issue, that is the way to go, IMO.
Sigurd
If a few R-core transformers can fix the current issue, that is the way to go, IMO.
Sigurd
syn08 said:
Sigurd Ruschkow said:
Interesting, why would you hesitate? You think it's a hazard for the cartridge or the sound will be affected?
I've read a lot about input caps as being a big nono in MC/MM, and though high end preamps are using such a cap? Where does the whole "caps in the signal path" story, promoted by the "high end experts", fit?
And even so, 47k input impedance and 10Hz rolloff requires a 0.33uF cap. But a 100ohm input impedance requires 15uF. Such a film cap (I suppose they don't use electrolytics) is an ideal antenna for picking up noise. 10ohm impedance requires 150uF! How are those expensive preamps addressing this problem?
Hesitate because I think it can be a hazard for the cartridge.
Well, there are several caps in series with the signal in any RC-based timing network. Small, though, and one can use even bigger than normal caps.
But the one to stop DC currents from flowing in the cartridge has to be large in value as you say, around 100u, so the ones I have seen are thus (smallish) electrolytic capacitors...
It would be very interesting to hear what say Clearaudio has to say about DC currents in their MC cartridges.
Sigurd
Well, there are several caps in series with the signal in any RC-based timing network. Small, though, and one can use even bigger than normal caps.
But the one to stop DC currents from flowing in the cartridge has to be large in value as you say, around 100u, so the ones I have seen are thus (smallish) electrolytic capacitors...
It would be very interesting to hear what say Clearaudio has to say about DC currents in their MC cartridges.
Sigurd
syn08 said:
Sigurd Ruschkow said:
So you think an electrolytic cap in series with the MC cartridge is ok... I wonder what the high end talibans have to say about
Now, if it's about the hazard created by that current, I would certainly forget about. The thinnest magnet wire I have ever heard about is gold, 10um diameter. 1uA through such a wire would render a current density of not more than 80mA/mm^2, which is absolutely negligible. Even a 1um dia wire will carry a current density of under 1A/mm^2, which is still far away from any adverse effects. OTOH, no cap will help in avoiding a current pulse through the MC cartridge, at least at power on and power off. And electrolytics may have leakages of that order of magnitude anyway.
No, you know that I do not like caps in the signal path when it is not really neccessary (like in a RIAA to create the time constants)
but I wanted to let you know that even very expensive MC RIAA amplifiers do sometimes use electrolytic caps in series with the cartridge.
One might even argue that the cap in a DC servo is in the signal path but they are sometimes necessary. Alas, I do use very good (PP or PS) caps for the DC servo caps.
Do you think that the noise increase with a servo of some kind could be decreased with roughly the same noise amount by aparallelling more input devices? Wouldn't that be an elegant solution?
Sigurd
but I wanted to let you know that even very expensive MC RIAA amplifiers do sometimes use electrolytic caps in series with the cartridge.
One might even argue that the cap in a DC servo is in the signal path but they are sometimes necessary. Alas, I do use very good (PP or PS) caps for the DC servo caps.
Do you think that the noise increase with a servo of some kind could be decreased with roughly the same noise amount by aparallelling more input devices? Wouldn't that be an elegant solution?
Sigurd
syn08 said:
So you think an electrolytic cap in series with the MC cartridge is ok... I wonder what the high end talibans have to say about
Now, if it's about the hazard created by that current, I would certainly forget about. The thinnest magnet wire I have ever heard about is gold, 10um diameter. 1uA through such a wire would render a current density of not more than 80mA/mm^2, which is absolutely negligible. Even a 1um dia wire will carry a current density of under 1A/mm^2, which is still far away from any adverse effects. OTOH, no cap will help in avoiding a current pulse through the MC cartridge, at least at power on and power off. And electrolytics may have leakages of that order of magnitude anyway.
AndrewT said:
Agreed, but Sigurd says 1uA through a MC can be dangerous?! If it's not, why using a cap at all?
Sigurd: could you name a high end MC preamp that is using an electrolytic at the input? If an electrolytic is used at the input, do you think it makes sense to use teflon, or whatever other exotic caps, anywhere else in the signal path? To me, no other cap could damage the sound worse than a non-linear, leaking, polarised, DA, etc... as an electrolytic.
Yes, I know, the power is in the uW range. I am not saying that an MC cartridge will change sonically or become broken, but I am trying to figure out what we are dealing with here, and what levels of current can we accept (over time and temperature).
The Ortofon has 100uV at 5 cm/s, right, Syn08? Would that be considered "0 dB", and then the minimal signal level from the Ortofon at -60 dB (dynamic range of LP:s) is
0.1uV?
Sigurd
The Ortofon has 100uV at 5 cm/s, right, Syn08? Would that be considered "0 dB", and then the minimal signal level from the Ortofon at -60 dB (dynamic range of LP:s) is
0.1uV?
Sigurd
AndrewT said:
Until I kknow more about this, yes, I do feel uncomfortable with DC currents through a MC cartridge. I am not saying that it is hazardous nor sonically bad, but I would like to understand more about what we are dealing with here.
I am using JFETs for input stage right now, but it seems that the future is not not JFETbased even though today there is easy to get the complementary Toshiba JFETs, the SJ74 is discontinued,
so the more we know about DC currents through a MC cartridge, the bestter, no?
Well, this high end MC preamp did in no way sound perfect. Sure, it is expensive (10k Euro) but the sound can be better for less money. Maybe the electrolytic cap degraded the sound or was it maybe th eopen loop design that made it sound less than "perfect"? I do not know. Compared to a 5K USD OPamp based RIAA, highs were better on the more expensive RIAA amplifier.
Yeah, I dislike elektrolytics like you do, too. I use Teflon and polystyrene whenever pissible. Or film popypropylene.
Syn08, I am trying to inspire you to see the challenge in getting rid of the DC current through the cartridge.
Sigurd
I am using JFETs for input stage right now, but it seems that the future is not not JFETbased even though today there is easy to get the complementary Toshiba JFETs, the SJ74 is discontinued,
so the more we know about DC currents through a MC cartridge, the bestter, no?
Well, this high end MC preamp did in no way sound perfect. Sure, it is expensive (10k Euro) but the sound can be better for less money. Maybe the electrolytic cap degraded the sound or was it maybe th eopen loop design that made it sound less than "perfect"? I do not know. Compared to a 5K USD OPamp based RIAA, highs were better on the more expensive RIAA amplifier.
Yeah, I dislike elektrolytics like you do, too. I use Teflon and polystyrene whenever pissible. Or film popypropylene.
Syn08, I am trying to inspire you to see the challenge in getting rid of the DC current through the cartridge.
Sigurd
syn08 said:
Good points, Wavebourn.
Yes, we could use a transformer. There are a few from Lundahl and Jensen that looks interesting. The price might be too high for some, though and or the sonics might not be what one is looking for. It also is not easy to use a transformer that fits all MC cartridges impedance-wise.
How high would you estimate the charge current to be?
Sigurd
Yes, we could use a transformer. There are a few from Lundahl and Jensen that looks interesting. The price might be too high for some, though and or the sonics might not be what one is looking for. It also is not easy to use a transformer that fits all MC cartridges impedance-wise.
How high would you estimate the charge current to be?
Sigurd
Wavebourn said:
If you decide to go with transformer, why do we need multiple low rbb pairs at all? 6J52P would be more preferable.
If to use a cap, are you sure it's charge current can't damage a $1,000 cartridge, especially if the cap is low loss one?
No, you are not.
You are completely ignoring your own point, that is that even high-end stuff put an elco as DC-blocker inbetween, although an elco has considerable leakage. Leakage is nothing else than DC-current.
Also what does 'getting rid' mean?
There will always be a DC-current (even with jfets, see gate leakage), be it pA (picoAmp) or whatever. The point is to know when you can completely forget about it.
By the way, a rough estimate on neglible DC* would be IMHO that the dc-current is less than 1% of the nominal output current (shorted outputs); for a 1mV cartridge with 10R resistance that would make 1 muA.
Have fun, Hannes
*neglible in a signal sense (offsetting the armature etc), heating due to current I cannot imagine to be much of a concern as other factors will deteriorate the signal first.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.