I've recently gotten my hands on an Eico 2036 integrated amplifier. I've stripped it down to the basic amplifier circuit and have a few questions. I haven't been able to locate a schematic so I pieced one together, attached. I have not replaced any components as of yet, but I have used it the last several nights and it sounds quite good.
1. The amp has what I have found to be called a PEC (printed electric circuit). Without a schematic, I am guessing as to what it does, my assumption is shown on my schematic. One thing that struck me as odd was a 10k resistor on the grid of one output tube but not the other, maybe it is within the PEC for the other tube?
2. I intend to replace the PEC with individual components. I can't locate 87k resistors easily, which would be better- to go up or down in value? I understand they should be closely matched. I have some 47k resistors I was thinking of, too big of a difference? I'm not quite sure how to draw a loadline for this particular case.
3. Power transformer runs rather warm. Other threads suggest that if you can't keep your hand on it that it is too warm. I measured about 150°F after a couple hours of playing. I am a little worried because I have already lightened the load by not installed 4 tubes that would be in the complete amp (phono and tone sections). Maybe its not an issue? I will definitely replace filter caps before proceeding any further, maybe one or more sections is leaky and drawing some extra current. All tubes seem to be biased properly and drawing normal amount of current.
Any help with the above or a schematic is appreciated.
1. The amp has what I have found to be called a PEC (printed electric circuit). Without a schematic, I am guessing as to what it does, my assumption is shown on my schematic. One thing that struck me as odd was a 10k resistor on the grid of one output tube but not the other, maybe it is within the PEC for the other tube?
2. I intend to replace the PEC with individual components. I can't locate 87k resistors easily, which would be better- to go up or down in value? I understand they should be closely matched. I have some 47k resistors I was thinking of, too big of a difference? I'm not quite sure how to draw a loadline for this particular case.
3. Power transformer runs rather warm. Other threads suggest that if you can't keep your hand on it that it is too warm. I measured about 150°F after a couple hours of playing. I am a little worried because I have already lightened the load by not installed 4 tubes that would be in the complete amp (phono and tone sections). Maybe its not an issue? I will definitely replace filter caps before proceeding any further, maybe one or more sections is leaky and drawing some extra current. All tubes seem to be biased properly and drawing normal amount of current.
Any help with the above or a schematic is appreciated.
1 probably
2 47K will result in nearly twice the current being drawn by the tube. Without having power supply voltages it is difficult to estimate the change in dissipation, however I would suggest 86K for the replacement as that is the next lower standard value.
3 line voltage today can be up to 15% greater than when the amp was built. You might consider adding a thermistor in the AC Line input to the transformer. This will help with input surge as well as dropping a couple volts and aiding in lowering your B+ and more importantly heater voltage. Check your heater voltage to the tubes and see if it is within 5% of 12.6V (Less than 13.23V).
2 47K will result in nearly twice the current being drawn by the tube. Without having power supply voltages it is difficult to estimate the change in dissipation, however I would suggest 86K for the replacement as that is the next lower standard value.
3 line voltage today can be up to 15% greater than when the amp was built. You might consider adding a thermistor in the AC Line input to the transformer. This will help with input surge as well as dropping a couple volts and aiding in lowering your B+ and more importantly heater voltage. Check your heater voltage to the tubes and see if it is within 5% of 12.6V (Less than 13.23V).
Hi, The 10K is a "grid stopper". This is used in conjunction with the tube's internal capacitance to form a low pass filter (or high pass cut if you prefer). It normally is done to prevent oscillation. For the 87K resistors I would series a couple to get close to the value. two 33K and one 22K would be real close. Many other combinations are possible.
Copy the ST-70 driver stage and you'll be happy....I had to do the same after a PEC burned out. Nice improvement....
[1] However you derived the capacitance values for the printed circuit module … those resistances in there will have fung'd up the cap values.
[2] Ever heard of "mouser.com"? Precision resistors are like 13 cents apiece in ½ watt metal-film units. 75 K, ½ W, 1% Vishay is 13 cents. So yes, you can find 'em. These are even "stocked", so there is little-to-no wait time.
[2b] Then there's those 377 K (or whatever they are) resistors … which turn out to be just as important in their relative precision as the 87 K units. They are in the signal path as shunts, in a “A/C voltage divider” sense. However, their absolute value isn't as important as their precision-in-matching.
[2c] The Gimp's advice regarding "standard values" replacement is more important than it may seem on the surface. His advice is like the tip of an iceberg: there's 11× more 'berg' below the waves! Few people know, fewer accept and even fewer embrace the notion that absolute-value precision in 99% of circuits is a false ideal; relative precision is far more usual as a true reason for specification.
[2d] As a for-instance, the 165 Ω cathode self-bias resistor on the EL84 finals. It could realistically be anything from 150 to 180 ohms. Or even a wider range, depending on the design point of the finals and the actual high voltage power supply.
[2×10] Again, reflecting on The Gimp's sage words: you can always "fiddle" with resistance values by putting a couple in series to get a less commonly available value. The 377 K resistors are particularly thus. Remember - so long as they're matched - anything from about 330 KΩ to 420 KΩ will suffice in the position.
[3] Yet again echoing The Gimp, line voltage is something which has changed substantially over the years, and between countries. The idea of using a thermistor is OK, as it acts like a soft-conduction voltage dropper, and is built to dissipate some power. The "other side" is also a potential target. Just change out the 165 Ω cathode bias resistor on the finals for a 200 Ω resistor; it will drop the current borne by the finals, which "down wind" will affect all the rest of power consumption (at the expense of diminishing output a bit).
[3b] Giving consideration to your observation that the transformer runs hot… another way to reduce the whole HV section is to insert a robust modest-value resistor between the HV winding center-tap and ground. You of course need to figure out the total HV amperage being drawn, but since you have that 50 Ω resistor there on the hot-side, it makes a pretty good current-to-voltage shunt. Just be careful in measuring the voltage across it! Its live, and it can kill you.
[3c] the comment regarding measuring the A/C heater voltage is very good: its not like you really have to worry about a bit of heater over-voltage, because the remarkably non-linear temperature coefficient of tungsten heater wire ensures that the true dissipated wattage is much less different with a pretty wide range of heater voltage, than one might expect from an ordinary resistor … but it again makes a good measurement proxy to determine if your line voltage (in combination with the power transformer's primary) is significantly over spec.
Myself … I wouldn't be worried of the wall-socket over voltage unless it is egregiously so. I'd just drop the Hi-V side either with that resistor-in-series-with-the-center-tap, or, by changing the 50 Ω first high-tension resistor for something larger. (And higher power…)
GoatGuy
[2] Ever heard of "mouser.com"? Precision resistors are like 13 cents apiece in ½ watt metal-film units. 75 K, ½ W, 1% Vishay is 13 cents. So yes, you can find 'em. These are even "stocked", so there is little-to-no wait time.
[2b] Then there's those 377 K (or whatever they are) resistors … which turn out to be just as important in their relative precision as the 87 K units. They are in the signal path as shunts, in a “A/C voltage divider” sense. However, their absolute value isn't as important as their precision-in-matching.
[2c] The Gimp's advice regarding "standard values" replacement is more important than it may seem on the surface. His advice is like the tip of an iceberg: there's 11× more 'berg' below the waves! Few people know, fewer accept and even fewer embrace the notion that absolute-value precision in 99% of circuits is a false ideal; relative precision is far more usual as a true reason for specification.
[2d] As a for-instance, the 165 Ω cathode self-bias resistor on the EL84 finals. It could realistically be anything from 150 to 180 ohms. Or even a wider range, depending on the design point of the finals and the actual high voltage power supply.
[2×10] Again, reflecting on The Gimp's sage words: you can always "fiddle" with resistance values by putting a couple in series to get a less commonly available value. The 377 K resistors are particularly thus. Remember - so long as they're matched - anything from about 330 KΩ to 420 KΩ will suffice in the position.
[3] Yet again echoing The Gimp, line voltage is something which has changed substantially over the years, and between countries. The idea of using a thermistor is OK, as it acts like a soft-conduction voltage dropper, and is built to dissipate some power. The "other side" is also a potential target. Just change out the 165 Ω cathode bias resistor on the finals for a 200 Ω resistor; it will drop the current borne by the finals, which "down wind" will affect all the rest of power consumption (at the expense of diminishing output a bit).
[3b] Giving consideration to your observation that the transformer runs hot… another way to reduce the whole HV section is to insert a robust modest-value resistor between the HV winding center-tap and ground. You of course need to figure out the total HV amperage being drawn, but since you have that 50 Ω resistor there on the hot-side, it makes a pretty good current-to-voltage shunt. Just be careful in measuring the voltage across it! Its live, and it can kill you.
[3c] the comment regarding measuring the A/C heater voltage is very good: its not like you really have to worry about a bit of heater over-voltage, because the remarkably non-linear temperature coefficient of tungsten heater wire ensures that the true dissipated wattage is much less different with a pretty wide range of heater voltage, than one might expect from an ordinary resistor … but it again makes a good measurement proxy to determine if your line voltage (in combination with the power transformer's primary) is significantly over spec.
Myself … I wouldn't be worried of the wall-socket over voltage unless it is egregiously so. I'd just drop the Hi-V side either with that resistor-in-series-with-the-center-tap, or, by changing the 50 Ω first high-tension resistor for something larger. (And higher power…)
GoatGuy
Thank you all for the assistance so far.
Mr GoatGuy, yes I get most of my parts on Mouser.com. Maybe I'm doing something wrong with the search, but the only result that comes up for 87k is a 1/20 watt unit which I doubt would be suitable, if not for power dissapation the voltage rating. However I'm sure I can find something else that is in the ball park.
Everything else seems fairly reasonable, and I appreciate your time taken to explain everything.
Mr GoatGuy, yes I get most of my parts on Mouser.com. Maybe I'm doing something wrong with the search, but the only result that comes up for 87k is a 1/20 watt unit which I doubt would be suitable, if not for power dissapation the voltage rating. However I'm sure I can find something else that is in the ball park.
Everything else seems fairly reasonable, and I appreciate your time taken to explain everything.
MF1/2CC8872F KOA Speer | Mouser
As stated it doesn't have to be exactly 87k, most likely the original carbon composition were +/- 5% in value.
EDIT: Noticed they were on order. But check these out, little more money but your only buying two🙂 http://www.mouser.com/ProductDetail...=sGAEpiMZZMu61qfTUdNhG3gu04FpCMicVmdauPR03u4=
As stated it doesn't have to be exactly 87k, most likely the original carbon composition were +/- 5% in value.
EDIT: Noticed they were on order. But check these out, little more money but your only buying two🙂 http://www.mouser.com/ProductDetail...=sGAEpiMZZMu61qfTUdNhG3gu04FpCMicVmdauPR03u4=
Last edited:
594-SFR2500008662FR5 = mouser number. 1%, half watt, 86.6kΩ
If you take the gist of what I wrote above, it is not the absolute value (87K) which matters, but the closeness-of-matching somewhere in the right absolute-value range. The difference alone is less than 0.5% (between 86.6kΩ and 87.0kΩ)
TO FIND THIS PART, I went to mouser and used the filters.
power rating = 0.25 to 2 watts
resistance = 75kΩ to 90kΩ
termination style = axial
tolerance = 0.25% to 1%
stocked = YES (checkboxed)
Came up with over 50 alternatives that are stocked. There are at least 3 that are "86.6kΩ".
Here's your 377kΩ replacement: 594–5043ED374K0F Actual ohmage is 374kΩ. Again … absolute match carries no magic in and of itself. It is the relative match-of-precision that counts. $0.12 apiece. Buy 10!
By comparison, you're going to be vexed by replacing the capacitors with anything that is matched. Luckily, the "purpose" of the caps is to block DC, and to be pretty good as low-Z passers-of-signal to the grids of the finals. So, don't sweat it. Just get some nice caps with not-to-expensive costs, and good reputations. And plenty of DC blocking voltage!
Myself, I would very likely insert a 10kΩ grid-stop resistor in the other EL84 grid line. Its omission is of course almost certainly of no negative consequence, but the 10kΩ one on the "A" side suggests they had oscillation problems without it, and its presence squashed that problem. Economics, and not original design, most likely resulted in the single-resistor being there.
Good luck. Remember: it is not about absolutely accurate value-matches, but for critical sections (such as this phase-splitter), the relative match of components. And, sufficient power dissipation. There's going to be over 300 VDC of the power supply … and the 87kΩ resistors will each take (in worst case) ½ of that, or 150 VDC. Remembering that P = E²/R then P = 150²/87000 = 0.26 watts. Actual power loading will be lower because (of course) the tube will be taking about ½ to ⅓ itself.
GoatGuy
Last edited:
I suggest that the OP take a long look at the Dyna ST35 schematic, given the same signal tube complement.
My understanding of this thread is that the OP wants to use the power O/P portion of the EICO original. The EICO O/P "iron" will not support UL mode "finals", but pure pentode mode can be made to work very well by regulating g2 B+. Take a Chinese/American restaurant attitude and pick 1 from column "A" and 1 from column "B". Both EICO and Dyna provide good clues as to what the ultimate setup should be. As this is clearly not an "exact" restoration project, shoot for the maximal result obtainable from what is on hand.
FWIW, freed up socket positions could be used to hold an 0A2/0B2 gas regulator stack that provides O/P tube g2 B+. The extra heater current allows for use of a 12BH7 to provide better "concertina" phase splitters, than the 12AU7 sections of the 7247s. Use a 12AX7 to provide the voltage amplifiers. Healthy used old stock 7247s should be marketable.
My understanding of this thread is that the OP wants to use the power O/P portion of the EICO original. The EICO O/P "iron" will not support UL mode "finals", but pure pentode mode can be made to work very well by regulating g2 B+. Take a Chinese/American restaurant attitude and pick 1 from column "A" and 1 from column "B". Both EICO and Dyna provide good clues as to what the ultimate setup should be. As this is clearly not an "exact" restoration project, shoot for the maximal result obtainable from what is on hand.
FWIW, freed up socket positions could be used to hold an 0A2/0B2 gas regulator stack that provides O/P tube g2 B+. The extra heater current allows for use of a 12BH7 to provide better "concertina" phase splitters, than the 12AU7 sections of the 7247s. Use a 12AX7 to provide the voltage amplifiers. Healthy used old stock 7247s should be marketable.
Last edited:
AS this amplifier was obviously quite old, the component values would have been at best E12 series. So 87k was probably 82k etc.
- Status
- Not open for further replies.