Tweet
Garnet Mach 5 from 1969
This is my first post ever anywhere. Forgive my small novel. I have a Garnet Mach 5 212 combo. It is PTP wiring...with two 12AX7s, one 6SN7, & two 6V6s. First 12AX7 is 2-state preamp which feeds the 6SN7 setup as a Cathodyne inverter which precedes the PP output. The other 12AX7 operates a phase-shift oscillator circuit (trem) and feeds into the output tubes. Controls are on/off switch, vol, bass, treble, tremolo on/off, depth, and speed. Twin input jacks, pilot light, and 3A fuse.
Tremolo has never worked. The 40-40-20-20 cap was replaced ~4 yrs ago. I put it thru some aggressive playing, ending with some hard drop-D and volume dimed. Within a few mins, the amp emitted a sound like speaker voice coils being shredded. I turned it off quickly. Speakers checked out fine with another amp. All 12AX7 and 6V6 tubes were replaced one at a time with known working tubes...and same shredding noise remained. All tubes also worked fine as replacements in the other working amp (6SN7 unique so not able to replace). At that point I assumed it was either the 6SN7, a cap, or one of the XFMRs. The amp has set idle since.
This past weekend I decided to see if I could isolate the cause. I plugged the amp into a bulb style current limiter with all tubes installed. After the initial faint glow at turn (filter in-rush) the bulb went unlit. After a few seconds the 6V6s heaters warmed up and current started to flow in the output and the bulb would then start to light up, although only about 1/3 of its brightness. I quickly turned off the amp. I drained the filter caps, then I setup my Fluke 87 to measure the 6V6 plate. At turn on the plate quickly comes up to ~400V. As soon as the 6V6s started to conduct the plate was pulled down to just a few volts (bias of 180ohm feeding both 6V6s).
I removed all tubes and tried the same measurement at each socket using only a single tube each trial. Only the output sockets exhibited the high current draw. The phenomenon seem to respond like a tube shorted internally. So I installed some working 6L6s, and got the same result with a little more bulb brightness. I checked out the output xfmr by inputting a sine wave into the secondary and measuring the primary. The voltage checked out exactly 50-50 on each plate compared to CT on primary. I attempted to measure the power xfmr laminates for shorting and did not find any shorts, although I'm not sure I'm doing it right. So I completely disconnected the power xfmr and connected up a brand new tranny. Again no change. All things seem to point to the socket having a short which I just couldn't believe. So to isolate the socket I diconnected the grid on the one 6V6 socket and the heater wires running in parallel to all other sockets. I put in new diodes, bypased the multi-can filter, put in a new single filter cap, and basically hooked up the one output socket like a stand-alone pentode with 200K resistor plate load. I did keep the original 180ohm cathode bias (other 6V6 removed from socket). The 180 ohm was measured as 177 ohms. I installed a 6L6 to handle the high plate voltage. At power up the plate quickly went up to ~520 volts waiting for the heaters to warm up. After a few seconds the plate started to drop down to ~200 volts rapidly, then raised back up to ~300 volts and remained. Given the single filter cap of 40uF, the drop was more than I expected but it settled out reasonably. The current limit bulb remained "off". So this seem to confirm the socket is not the problem. Plus the old diodes measured 0.5V on the Fluke so they seem fine too.
At this point I was out of technique and knowledge
, and decided I needed help. Any ideas?
Tremolo has never worked. The 40-40-20-20 cap was replaced ~4 yrs ago. I put it thru some aggressive playing, ending with some hard drop-D and volume dimed. Within a few mins, the amp emitted a sound like speaker voice coils being shredded. I turned it off quickly. Speakers checked out fine with another amp. All 12AX7 and 6V6 tubes were replaced one at a time with known working tubes...and same shredding noise remained. All tubes also worked fine as replacements in the other working amp (6SN7 unique so not able to replace). At that point I assumed it was either the 6SN7, a cap, or one of the XFMRs. The amp has set idle since.
This past weekend I decided to see if I could isolate the cause. I plugged the amp into a bulb style current limiter with all tubes installed. After the initial faint glow at turn (filter in-rush) the bulb went unlit. After a few seconds the 6V6s heaters warmed up and current started to flow in the output and the bulb would then start to light up, although only about 1/3 of its brightness. I quickly turned off the amp. I drained the filter caps, then I setup my Fluke 87 to measure the 6V6 plate. At turn on the plate quickly comes up to ~400V. As soon as the 6V6s started to conduct the plate was pulled down to just a few volts (bias of 180ohm feeding both 6V6s).
I removed all tubes and tried the same measurement at each socket using only a single tube each trial. Only the output sockets exhibited the high current draw. The phenomenon seem to respond like a tube shorted internally. So I installed some working 6L6s, and got the same result with a little more bulb brightness. I checked out the output xfmr by inputting a sine wave into the secondary and measuring the primary. The voltage checked out exactly 50-50 on each plate compared to CT on primary. I attempted to measure the power xfmr laminates for shorting and did not find any shorts, although I'm not sure I'm doing it right. So I completely disconnected the power xfmr and connected up a brand new tranny. Again no change. All things seem to point to the socket having a short which I just couldn't believe. So to isolate the socket I diconnected the grid on the one 6V6 socket and the heater wires running in parallel to all other sockets. I put in new diodes, bypased the multi-can filter, put in a new single filter cap, and basically hooked up the one output socket like a stand-alone pentode with 200K resistor plate load. I did keep the original 180ohm cathode bias (other 6V6 removed from socket). The 180 ohm was measured as 177 ohms. I installed a 6L6 to handle the high plate voltage. At power up the plate quickly went up to ~520 volts waiting for the heaters to warm up. After a few seconds the plate started to drop down to ~200 volts rapidly, then raised back up to ~300 volts and remained. Given the single filter cap of 40uF, the drop was more than I expected but it settled out reasonably. The current limit bulb remained "off". So this seem to confirm the socket is not the problem. Plus the old diodes measured 0.5V on the Fluke so they seem fine too.
At this point I was out of technique and knowledge
, and decided I needed help. Any ideas?
Comment