• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

ECC83 input circuitry

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I am at last getting around to building a valve amp based on the original Tripletone Hi-Fi Major that I bought back in 1967 and used for many years until my audio system became transistorised. Oh for the benefit of hindsight!

The original amp was dismantled without any details being recorded. Fortunately however, I kept the transformers and valves. After assistance from a number of people I now have most of the circuitry sorted out, sufficient at least to make a start on the reconstruction. Unfortunately I am without any information on how I should set up the input to the grid of the first stage which is half an ECC83. I want to get the amp as close as possible to the original spec which was for a ceramic cartridge with an 80mV sensitivity at 2M ohm impedance.

Has anyone any idea of the component values I will need?
 
An ECC83 data sheet will help.

Granted it would if I knew how to interpret the data. I'm afraid audio is a bit low frequency for me as I'm used to constructing valve RF projects. I think in this case I will need to experiment to get the values somewhere near to present the correct impedance to the cartridge and then I can work on the sensitivity aspect.

I posed the question in case anyone could provide an answer based on experience so I had a starting point to work from.
 
If you understand valve RF then much (but not all) of valve audio should be easy. Going the other way is more difficult.

No need to experiment. For 2M input impedance you just need a 2M grid resistor - but check the data sheet to see if this satisfies the maximum value; I think it will be OK. For 80mV sensitivity you will need to do some gain calculations. I'm not sure why you want to match a ceramic cartridge, as very few people use them nowadays and the input will be too sensitive for any other flat source.

As SY says, data sheets give suggested circuits. That is why I directed you to them.
 
I want to get the amp as close as possible to the original spec which was for a ceramic cartridge with an 80mV sensitivity at 2M ohm impedance.

This is unrealistic. The spec sheet for the ECC83 (12AX7) gives:

Cgk= 1.6pF
Crt= 1.6pF

Given that this is a high gain type, expect a gain of ~60V/V. (It will be even higher if you use an active plate load -- a really good idea for a type with rp= 90K (nom). ) Though the capacitance doesn't look like much, the Cmiller will be fierce:

Cmiller= Crt(1 + Av)
Cmiller= 1.6(1 + 60)= 97.6pF

Ci= 1.6 + 97.6= 99.2pF (Call that 100pF)

w= 1 / RC
w= 1 / (2.0E6 X 1.0E-12)= 5000 rad/sec
f= 5000 / 2pi= 795.77Hz

Add in more capacitance from lead-in's and that drops even more. You simply will have no highs at all. The only way you're gonna get that 2MEG of Zi is to buffer the input to the ECC83. You could use a source follower/emitter follower with an active tail load, or an op-amp voltage follower. If you want to avoid the SS, then a 6J5 with an active tail load ahead of the ECC83 will get you your 2MEG of Zi, avoid the capacitance problem, and a 6J5 with an active tail load produces almost no THD, and is quite transparent sonically.

If you don't want to do that, then you need to consider ditching the ECC83 altogether, and use either a small signal pentode like a 6AU6 (not exotic and "sexy" -- just a reliable old workhorse that does well for almost anything: AF, RF, CCS duty) or a cascode (6BQ7s work very nicely for this purpose) to avoid that Cmiller problem.

Has anyone any idea of the component values I will need?

ECC83 spec sheet from Frank's here

I'm not sure why you want to match a ceramic cartridge, as very few people use them nowadays and the input will be too sensitive for any other flat source.

Not true: lots of cheap (and not so cheap) record players use ceramic carts. I got one, and it came complete with an absolutely hideous chip amp. That damn thing sounded like
An externally hosted image should be here but it was not working when we last tested it.
. On top of that, the damn thing wasn't even stable! rewired the whole thing so's it could connect to the really decent VT amps. Sounds pretty good now, actually.
 
Last edited:
An interesting conclusion and one that is at loggerheads with the original spec for the amp which claimed 80mV into 2M over the range 30 - 20000Hz +/- 3dB or something like that. I used it with an Acos GP94/1 cartridge and had no issues at all. It sounded okay and contrary to what some might expect, didn't trash my vinyl. I recall tracking around 2 to 2.5g if my memory serves me right and the turntable was a Garrard SP25 Mk 1.

My original aim was to rebuild the amp as close to the original spec as possible. I have established the majority of the circuit by exchanging information with several people, but no-one has the same version of the amp as the one that I owned, so certain elements such as the input circuitry are lacking. I guess I keep looking.....
 
Last edited:
A ceramic cartridge behaves like a capacitor, so it can cope with a bit of Miller capacitance. Other high impedance sources might not cope, though.

Miles Prower said:
Not true: lots of cheap (and not so cheap) record players use ceramic carts.
Sorry. I assumed that anyone buying a record player nowadays would be interested in sound quality and preserving the grooves from stylus damage.

newmiac said:
An interesting conclusion and one that is at loggerheads with the original spec for the amp which claimed 80mV into 2M over the range 30 - 20000Hz +/- 3dB or something like that.
The spec was probably met when driven from a voltage source i.e. low impedance.
 
The spec was probably met when driven from a voltage source i.e. low impedance.[/QUOTE]

The amp had three inputs - ceramic, radio and tape. The radio and tape were typically 100k impedance if I recall correctly. I just wonder whether the 2M impedance was achieved by adding passive components in line to make the cartridge see 2M. There was definitely no difference in spec between the inputs.

The amp line up was ECC83 - ECC83 - 2 x EL84. It was all packaged into a steel cabinet about 11 x 7 x 3.5 inch with mesh grilles top and bottom. It had a clear perspex fascia painted green and white on its back. Three tone controls of course. The EZ81 rectifier and output valves were mounted with the transformers on a steel plate that was mounted vertically and ran left to right across the middle of the amp. The 83's were on a PCB in the front section. The design was an evolution from a more conventionally designed series of amps that were standard chassis designs and initially ran 2 x 6V6 making the change to EL84s in the early 60's.

I have only ever seen one of these amps for sale on an auction site and it went for mega-bucks. Unfortunately there was no way of contacting the purchaser and I had missed the auction by a couple of days. :(
 
Granted it would if I knew how to interpret the data. I'm afraid audio is a bit low frequency for me as I'm used to constructing valve RF projects. I think in this case I will need to experiment to get the values somewhere near to present the correct impedance to the cartridge and then I can work on the sensitivity aspect.

No, your tube with it's required 1M resistor from grid to ground will present a 1M impedance to the cartridge. The goal is not to load the cartridge, You are concerned with getting the max voltage swing not with maximum power transfer. That is one difference between audio and RF. It is very rare to try and match impedance in audio. The only case I can think of is when driving a loudspeaker (and in this case we DO want maximum power transfer)

But far more importantly a phono preamp need to apply the RIAA compensation curve. Those old vinyl records used a form of compression that you need to back out when you play them. Best to build from a knon working preamp design. Some are very simple, others not so simple and there must be 1,000 to choose from.

Tell us your budget, that is how many triode sections, you have to play with. Someone can recommend a schematic.
 
a) No, your tube with it's required 1M resistor from grid to ground will present a 1M impedance to the cartridge.

b) But far more importantly a phono preamp need to apply the RIAA compensation curve.

a) Yes, I know that putting 1M from grid to ground to will present 1M. Logically putting another 1M from the input to the grid will present 2M to the cartridge, but is this the correct approach? That is basically what I am asking.

b) The amp doesn't need RIAA compensation as the cartridge is ceramic not magnetic.

Cost is not an issue. I am trying to recreate an amplifier I owned in the late 60's / early 70's and keep as close to the original as possible, but as it stands I only have 95% of the circuitry established. The input arrangement is one issue and the values used in the negative feedback arrangement is the other.
 
regardless of the cart type, it still needs correction. The frequencies being corrected are those inherent in the grooves of the vinyl from its original cutting.

Your logic on cart loading is also flawed. The 2M load is as seen by the cart independent of the grid ie ACROSS (in shunt with) the cart output, not in series with it.
 
The 2M load is as seen by the cart independent of the grid ie ACROSS (in shunt with) the cart output, not in series with it.

I don't think that you understood what I meant. I was suggesting 1M from the input socket "live" connection to the grid and 1M from the grid to ground. The cartridge goes across the input socket "live" connection to ground, therefore the two resistors in series are across the cartridge terminals.

As for what causes the need for RIAA correction, I'll leave others to argue whether it is the cartridge or the vinyl. :)
 
OK, get you now on the cart load resistors. Of course, it means you are shunting half the signal away from the grid - not a great idea when there is so little signal in the first instance...

The other isn't an argument as such since I made a statement of fact. The RIAA curve is inherent in the groove of the record, independent of the cart type.
 
I think that the output from a ceramic cartridge is typically around 250mV, so it is not such an issue losing some across the resistor.

Apologies for the poor wording of the second part of my last post - I wasn't implying that there was an arguemnet to be had. :) Anyway, dragging information from the depths of my memory (after all it is a Friday!), I seem to recall that as you say the RIAA is an inherent characteristic of the record and since a correctly loaded ceramic cartridge exhibits a response that converts the response of the record to a flat one (or something close), there is no need for the electronics to supply the equalisation.
 
That 1M dropping resistor should make for a very noisy circuit.

Indeed, a ceramic cartridge has an output proportional to displacement, not velocity, so the cartridge output will have the same slope as RIAA. But... RIAA has an extra zero and pole at 318us and 75us, respectively, so the resulting frequency response will not be flat.
 
There are two aspects to RIAA, as usually applied to a magnetic cartridge. The generally falling response with frequency is compensating for the rising response with frequency of a magnetic cartridge, as this senses the velocity of the groove not the displacement. For a ceramic cartridge this correction is not needed, as it is a displacement transducer.

RIAA also has a 12dB mid-range shelf, from 500Hz to 2kHz (approx). This is needed for a ceramic cartridge, but can be omitted in cheaper record players. Omitting this give a rather muffled sound, but speaker and cabinet resonances can hide this!

BTW when driving a loudspeaker we do not want maximum power transfer, but a compromise between high efficiency and low distortion - genuine impedance matching is very rare. In RF we usually need impedances to match (or complex conjugate). In audio we need appropriate impedances, not matching.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.