TUBE Facts and Info by MartyStrat54
- Thread starter MartyStrat54
- Start date
This site may earn a commission from merchant affiliate links like Ebay, Amazon, and others.
MartyStrat54
Well-Known Member
I believe Mickey answered your question in your thread.
That would explain why my old JCM900 was the closest sounding amp I've ever heard to my AFD100? That made no sense eh? Nvm
Peace
bassistbobby
New Member
- Joined
- Sep 19, 2008
- Messages
- 83
- Reaction score
- 14
This is a GREAT article. Putting it all here is a great gift to all who are into tube amp ownership. This goes to the important subject of how to really enjoy your amp, not to just use it.
Being a JCM900 HGDR owner, I would just add that, in my own experience, changing my phase inverter tube, which is V3 in a JCM 900, to a long or medium plate tube, caused one of the biggest differences in tone I had come across in my own process of rolling my preamp tubes. I also discovered using a V2 tube with slightly higher emissions than my V1 tube, tended to smooth out the gain.
I have never read anything by anybody who says they've had the same experience, or who agrees with me. But this works for me every time.
Anyway,.... GREAT article Marty. Where were you back in '06 when I first bought my Marshall...? LOL. This type of knowledge back then would have saved me a lot of time, money, and grief.
Being a JCM900 HGDR owner, I would just add that, in my own experience, changing my phase inverter tube, which is V3 in a JCM 900, to a long or medium plate tube, caused one of the biggest differences in tone I had come across in my own process of rolling my preamp tubes. I also discovered using a V2 tube with slightly higher emissions than my V1 tube, tended to smooth out the gain.
I have never read anything by anybody who says they've had the same experience, or who agrees with me. But this works for me every time.
Anyway,.... GREAT article Marty. Where were you back in '06 when I first bought my Marshall...? LOL. This type of knowledge back then would have saved me a lot of time, money, and grief.
Metroman
Well-Known Member
- Joined
- Apr 29, 2014
- Messages
- 959
- Reaction score
- 859
I bought a Metro amps Limited Edition GMP45 Limited Edition. It has NOS GEC KT66's and Modern Mullard Preamp and Rectifier tubes.
I replaced the GZ34 with a 1961 NOS Mullard Fat Base.
Ive got a bunch of tubes and I have 14 NOS Mullard 10M Gold pins and am wondering if 3 of these would be the best choice for the amp?
I also have 11 NOS Telefunkens I haven't tried yet and I know the only way to find out is to try them.
Should I put a more transparent tube in v2 and v3 ? and Keep the 10M in v1 ? Or 3 10M's?
Im going towards the 3 10M's. Any advantage/disadvantage running the Midrangy Mullard 10M and transparent Telefunkens together ?
I replaced the GZ34 with a 1961 NOS Mullard Fat Base.
Ive got a bunch of tubes and I have 14 NOS Mullard 10M Gold pins and am wondering if 3 of these would be the best choice for the amp?
I also have 11 NOS Telefunkens I haven't tried yet and I know the only way to find out is to try them.
Should I put a more transparent tube in v2 and v3 ? and Keep the 10M in v1 ? Or 3 10M's?
Im going towards the 3 10M's. Any advantage/disadvantage running the Midrangy Mullard 10M and transparent Telefunkens together ?
MartyStrat54
Well-Known Member
The best thing is to roll them.
I would start out with a 10M in V1 and Tele's in V2 and V3.
I would start out with a 10M in V1 and Tele's in V2 and V3.
mickeydg5
Well-Known Member
- Joined
- Sep 30, 2011
- Messages
- 28,591
- Reaction score
- 16,367
Damn Metroman, you have some heavy hitter tubes in your collection.
No one mentioned smooth or ribbed Telefunken. From what I have gathered that usually makes some difference.
No one mentioned smooth or ribbed Telefunken. From what I have gathered that usually makes some difference.
yladrd61
Well-Known Member
I ended up going with a strong Blackburn I63 in V1, a Matsushita Black Plate in V2 and a CP Mullard in V3 on my GPM45..... My advice start with Martys suggestion and roll every preamp tube you have through each position and you will know which one you like best in each position.
yladrd61
Well-Known Member
I think I saw those 10M's on ebay I hope you didn't pay too much for them.... I also found that I like mismatched power tubes in this amp with the higher current draw in V4 
I have some Shuguang Treasure KT66-Z in mine with about a 5 ma mismatch.....
I have some Shuguang Treasure KT66-Z in mine with about a 5 ma mismatch.....
Metroman
Well-Known Member
- Joined
- Apr 29, 2014
- Messages
- 959
- Reaction score
- 859
Yeah I got it off Ebay. It tested over 93% and all parameters of it were really good.
I also sent it off and had it independently tested as it had a 30 day money back guarantee.
It tested as advertised.
I already had 2 of the Standard Pin Mullard 12AX7 10Ms and bought this 1 because they were originally tested at the factory for tight tolerances and microphonics. And the tubes are dead quiet.
Well worth the money and I get it cheaper ( Still Expensive ) than Ive seen anyone else have them for. both of the standard pin 10M's test 86% and 89%
I also sent it off and had it independently tested as it had a 30 day money back guarantee.
It tested as advertised.
I already had 2 of the Standard Pin Mullard 12AX7 10Ms and bought this 1 because they were originally tested at the factory for tight tolerances and microphonics. And the tubes are dead quiet.
Well worth the money and I get it cheaper ( Still Expensive ) than Ive seen anyone else have them for. both of the standard pin 10M's test 86% and 89%
Magictwanger
New Member
- Joined
- Jun 13, 2013
- Messages
- 55
- Reaction score
- 18
Shuguang El34m is a good Mullard copy. They are also re-branded as Groove tubes El34m with better testing!
EADGBE
Well-Known Member
- Joined
- Dec 13, 2009
- Messages
- 953
- Reaction score
- 1,342
I'll look into them.
Last edited:
EADGBE
Well-Known Member
- Joined
- Dec 13, 2009
- Messages
- 953
- Reaction score
- 1,342
Jaymz E
Well-Known Member
- Joined
- Sep 11, 2012
- Messages
- 1,466
- Reaction score
- 592
I ran across some old slightly used Russian 6n14n/el84s from the 70s for $25.00 an octet and have been using them in my Orange AD30tc and my Peavey Blues Classic 1x15 and they sound better than the PM Chinese and the regular Sovtek el84s. I put them in these amps about 2 months ago and they are still going strong.
MartyStrat54
Well-Known Member
I sold those a lot. They are very good, especially the older ones.
wallythacker
New Member
- Joined
- Oct 27, 2014
- Messages
- 88
- Reaction score
- 35
My two music stores, (I alternate cities) are very poor in keeping stock. One store at least as a pair of Electro Harmonix 12AXZ. Everything else, except for some Mesa EL84 Boogie output tubes scattered haphazardly around a display unit, has to be ordered. I don't understand that.
Hi Folks New to the forum from South Africa and thought I would share some info concerning Pre-amp tubes....I can't take all credit for some of the info below, as I don't know who the original authors are, full credit and kudos goes them...............anyway so here goes:
Valve comparison: From an Guitar players perspective
Graph Interpretation
All 12AX7s have an amplification factor (µ) of 100 ("high-mu"). Ideally every 12AX7 should yield the same gain when placed in the same circuit. However, in reality there are comparative gain differences among 12AX7s.
This graph is based on the average gain and noise measured from different tube brands, using the same number of samples for each brand. It should not be assumed that every single 12AX7 from a specific brand will exactly match its place on this graph.
The gain rating is only a relative reference (e.g. a rating of "10" does not mean ten times louder than a "1").
Numbers in parenthesis represent the average gain rating over the entire audio frequency range (Low, Mid, and High)
Voltage Gain Rating Noise Rating
10 = High Gain
1 = Low Gain 10 = High Noise / Microphonics
1 = Low Noise / Microphonics
Further, Why do Tubes sound different?
"It’s the RP-value, NOT the MU-value that’s Important!
To many people, the most important characteristic of a vacuum tube is its Amplification Factor (mu). Unfortunately, this is not true.
Of the vacuum tubes’ three operating characteristics – Dynamic Plate Resistance (rp), Transconductance (gm) and Amplification Factor (mu) - two are explicit values and one is an implicit value. The gm and rp values are explicit because they come directly from the tubes’ actual operating voltages and currents. The mu value, however, is an implicit value that is mathematically-derived from the product of the rp and gm values:
mu = gm * rp
Gm and rp are almost (but not exactly) inversely related. That is, as gm increases with increasing plate current, rp proportionately decreases. This inverse relationship causes mu to be virtually a constant that typically varies less than 10-15%.
The vacuum tube operating characteristics gm, rp and mu are “open-circuit” values, that is, they are calculated with NO external loads applied. In the real world, however, tubes are operated in circuits having both input and output loads. It is the effects of these “loads” (especially upon the plate) that make rp more important than mu. Here’s why...
When resistances are connected in parallel (symbol “||”), their resulting resistance is always LESS than the smallest resistance value. For example, consider two resistances, A = 10K ohm and B = 100K ohm, connected in parallel:
R = (A||B) = (A*B)/(A+B)
R = (10K*100K)/(10K + 100K) = 9.09K ~ 9.1K ohms
The same thing happens when a tube is operated with a plate load resistor (RP), its rp value is “loaded” in the same fashion as resistance-B “loaded” resistance-A in the example above. For instance, operating a 12AX7 (rp = 62.5K ohm) with a 100K ohm plate load resistor produces a circuit gain of only 61.5, not 100!
Why? Because, the ‘effective’ plate resistance (rp’) is no longer 62.5K ohms – it’s 38% LESS – only 38.5K ohms:
rp’ = rp||RP
rp’ = (62.5K*100K)/(62.5K + 100K) = 38.46K ~ 38.5K ohms
...where:
rp = Tube dynamic plate resistance, ohms (12AX7: 62.5K ohm)
RP = Circuit plate load resistance, ohms (typically 100K ohm)
Now, when tube gm is multiplied times rp’ a much LOWER amplification factor (mu’) value is obtained because of the LOWER ‘effective’ rp’ value:
mu’ = gm * rp’
...for 12AX7 gm = 0.0016 A/V, then:
mu’ = (0.0016 * 38.5K) = 61.5
And, when the output loading of the NEXT circuit (tube, tone stack, etc.) is included, the effective plate load becomes even LOWER. For example, assume the 12AX7 plate load resistor RP = 100K ohm feeds its signal into a tone stack having a resistance value of Ro = 500K ohms. Now, rp’ becomes three loads in parallel:
rp’ = (rp||RP||Ro)
rp’ = 1/( 1/rp + 1/RP + 1/Ro )
rp’ = 1/( 1/62.5K + 1/100K + 1/500K ) = 35714.3 ~ 35.7K
mu’ = (0.0016 * 35.7K) = 57.1
Thus, while the “open-circuit” gain of a 12AX7 is 100 with no load, its gain in a “real world” circuit (tube with loads) will be nearer to 60 (ie: 57-62), depending upon RP and Ro.
SUMMARY: “Mu is what the tube (alone) is capable of...Mu’ is what you actually get (in a circuit).”
Additionally, just as circuit AMPLIFICATION is dependent upon rp, so also is circuit FREQUENCY response or BANDWIDTH (BW), ie:
BW = 1/(2*PI*R*C)
...where: R = rp||RP||Ro
So, whenever we “swap” tubes, it’s the rp value of the new tube interacting with the RP, Ro, R and C values of the existing circuit that causes the sometimes “subtle” changes in amplification (gain) and tone (frequency response) that we hear.
And, here’s why...
The rp’ values shown in Table 1 represent preamp tube operation with circuit load resistances of RP = 100K ohms and Ro = 500K ohms, ie:
rp’ = (rp||RP||Ro)
The illustrative bandwidth (BW) values shown in Table 1 were calculated using R = rp’ and an assumed constant fictitious capacitance value of C = 446µF, ie:
BW = 1/(2*PI*R*C)
where: R = rp’ and C = 446µF
Table 1 - Preamp tubes sorted by MU values.
As you can see, the greater the rp’ value, the lower the bandwidth! Which means that tubes with LOW rp values (and thus circuits with LOW rp’ values) will have greater BW and frequency response and proportionately less loss-of-gain (ie: mu’) than tubes with HIGH rp values. Table 2 re-orders the preamp tubes listing by their BW, in descending order.
Table 2 - Preamp tubes sorted by BW values.
Thus, substituting a 5751 for a 12AX7 (or ECC83, 7025, etc.) typically reduces circuit gain by 28% and slightly increases BW by 4%. Similarly, substituting a 12AY7 for a 12AX7 reduces circuit gain by almost half (45%) but nearly doubles the BW. And, remember, the greater the BW, the more “sparkle” and “chime” we’re likely to perceive.
Why only “....likely to perceive”? Because, changing one preamp tubes’ BW does not affect the BW of ANY of the subsequent tubes along the signal path. There might be more frequencies in the drive signal to the following tubes but they will not (usually) be able to pass along those added frequencies to the speaker(s) because their BWs have not changed.
That is ‘why’ swapping preamp tubes seemingly produces vague and subtle, instead of dramatic, tonal changes, while swapping output power tubes (the last tubes in the path) often produces VERY noticeable sonic changes.
...an "old" USN electronics proverb goes:
SHORT = infinite BW, because there's no R working with C to cause an RC rolloff.
OPEN = zero BW, because there's infiinite R working with C to cause immediate RC rolloff.
However, in the 'whole' amp, there's always the inductance ( L ) of the audio output transformer (OT) to consider to, which typically is *the* limiting factor for both LOW(f1)- and HIGH(f2)-cutoff frequencies!" But then again they are designed for a specific application which is always a delicate balance.
Hope you found the above interesting and useful ...... Attila
Valve comparison: From an Guitar players perspective
Graph Interpretation
All 12AX7s have an amplification factor (µ) of 100 ("high-mu"). Ideally every 12AX7 should yield the same gain when placed in the same circuit. However, in reality there are comparative gain differences among 12AX7s.
This graph is based on the average gain and noise measured from different tube brands, using the same number of samples for each brand. It should not be assumed that every single 12AX7 from a specific brand will exactly match its place on this graph.
The gain rating is only a relative reference (e.g. a rating of "10" does not mean ten times louder than a "1").
Numbers in parenthesis represent the average gain rating over the entire audio frequency range (Low, Mid, and High)
Voltage Gain Rating Noise Rating
10 = High Gain
1 = Low Gain 10 = High Noise / Microphonics
1 = Low Noise / Microphonics
Further, Why do Tubes sound different?
"It’s the RP-value, NOT the MU-value that’s Important!
To many people, the most important characteristic of a vacuum tube is its Amplification Factor (mu). Unfortunately, this is not true.
Of the vacuum tubes’ three operating characteristics – Dynamic Plate Resistance (rp), Transconductance (gm) and Amplification Factor (mu) - two are explicit values and one is an implicit value. The gm and rp values are explicit because they come directly from the tubes’ actual operating voltages and currents. The mu value, however, is an implicit value that is mathematically-derived from the product of the rp and gm values:
mu = gm * rp
Gm and rp are almost (but not exactly) inversely related. That is, as gm increases with increasing plate current, rp proportionately decreases. This inverse relationship causes mu to be virtually a constant that typically varies less than 10-15%.
The vacuum tube operating characteristics gm, rp and mu are “open-circuit” values, that is, they are calculated with NO external loads applied. In the real world, however, tubes are operated in circuits having both input and output loads. It is the effects of these “loads” (especially upon the plate) that make rp more important than mu. Here’s why...
When resistances are connected in parallel (symbol “||”), their resulting resistance is always LESS than the smallest resistance value. For example, consider two resistances, A = 10K ohm and B = 100K ohm, connected in parallel:
R = (A||B) = (A*B)/(A+B)
R = (10K*100K)/(10K + 100K) = 9.09K ~ 9.1K ohms
The same thing happens when a tube is operated with a plate load resistor (RP), its rp value is “loaded” in the same fashion as resistance-B “loaded” resistance-A in the example above. For instance, operating a 12AX7 (rp = 62.5K ohm) with a 100K ohm plate load resistor produces a circuit gain of only 61.5, not 100!
Why? Because, the ‘effective’ plate resistance (rp’) is no longer 62.5K ohms – it’s 38% LESS – only 38.5K ohms:
rp’ = rp||RP
rp’ = (62.5K*100K)/(62.5K + 100K) = 38.46K ~ 38.5K ohms
...where:
rp = Tube dynamic plate resistance, ohms (12AX7: 62.5K ohm)
RP = Circuit plate load resistance, ohms (typically 100K ohm)
Now, when tube gm is multiplied times rp’ a much LOWER amplification factor (mu’) value is obtained because of the LOWER ‘effective’ rp’ value:
mu’ = gm * rp’
...for 12AX7 gm = 0.0016 A/V, then:
mu’ = (0.0016 * 38.5K) = 61.5
And, when the output loading of the NEXT circuit (tube, tone stack, etc.) is included, the effective plate load becomes even LOWER. For example, assume the 12AX7 plate load resistor RP = 100K ohm feeds its signal into a tone stack having a resistance value of Ro = 500K ohms. Now, rp’ becomes three loads in parallel:
rp’ = (rp||RP||Ro)
rp’ = 1/( 1/rp + 1/RP + 1/Ro )
rp’ = 1/( 1/62.5K + 1/100K + 1/500K ) = 35714.3 ~ 35.7K
mu’ = (0.0016 * 35.7K) = 57.1
Thus, while the “open-circuit” gain of a 12AX7 is 100 with no load, its gain in a “real world” circuit (tube with loads) will be nearer to 60 (ie: 57-62), depending upon RP and Ro.
SUMMARY: “Mu is what the tube (alone) is capable of...Mu’ is what you actually get (in a circuit).”
Additionally, just as circuit AMPLIFICATION is dependent upon rp, so also is circuit FREQUENCY response or BANDWIDTH (BW), ie:
BW = 1/(2*PI*R*C)
...where: R = rp||RP||Ro
So, whenever we “swap” tubes, it’s the rp value of the new tube interacting with the RP, Ro, R and C values of the existing circuit that causes the sometimes “subtle” changes in amplification (gain) and tone (frequency response) that we hear.
And, here’s why...
The rp’ values shown in Table 1 represent preamp tube operation with circuit load resistances of RP = 100K ohms and Ro = 500K ohms, ie:
rp’ = (rp||RP||Ro)
The illustrative bandwidth (BW) values shown in Table 1 were calculated using R = rp’ and an assumed constant fictitious capacitance value of C = 446µF, ie:
BW = 1/(2*PI*R*C)
where: R = rp’ and C = 446µF
Table 1 - Preamp tubes sorted by MU values.
As you can see, the greater the rp’ value, the lower the bandwidth! Which means that tubes with LOW rp values (and thus circuits with LOW rp’ values) will have greater BW and frequency response and proportionately less loss-of-gain (ie: mu’) than tubes with HIGH rp values. Table 2 re-orders the preamp tubes listing by their BW, in descending order.
Table 2 - Preamp tubes sorted by BW values.
Thus, substituting a 5751 for a 12AX7 (or ECC83, 7025, etc.) typically reduces circuit gain by 28% and slightly increases BW by 4%. Similarly, substituting a 12AY7 for a 12AX7 reduces circuit gain by almost half (45%) but nearly doubles the BW. And, remember, the greater the BW, the more “sparkle” and “chime” we’re likely to perceive.
Why only “....likely to perceive”? Because, changing one preamp tubes’ BW does not affect the BW of ANY of the subsequent tubes along the signal path. There might be more frequencies in the drive signal to the following tubes but they will not (usually) be able to pass along those added frequencies to the speaker(s) because their BWs have not changed.
That is ‘why’ swapping preamp tubes seemingly produces vague and subtle, instead of dramatic, tonal changes, while swapping output power tubes (the last tubes in the path) often produces VERY noticeable sonic changes.
...an "old" USN electronics proverb goes:
SHORT = infinite BW, because there's no R working with C to cause an RC rolloff.
OPEN = zero BW, because there's infiinite R working with C to cause immediate RC rolloff.
However, in the 'whole' amp, there's always the inductance ( L ) of the audio output transformer (OT) to consider to, which typically is *the* limiting factor for both LOW(f1)- and HIGH(f2)-cutoff frequencies!" But then again they are designed for a specific application which is always a delicate balance.
Hope you found the above interesting and useful ...... Attila
Last edited:
Joe6BG6GA
New Member
Greetings to all. I have found that a 12BZ7 gives a little more drive in 12AX7 driver/ phase inverter applications. However this type takes a little more heater current. You might like the change on sound.
