heathkit w-5m renovations [boondoggle edition] quantum vacuum deep dive

[sae-7125]

custom b+ circuit update.


the story thus far, including roadmap for electron navigation


closer view of the approximated active circuit on the bench

the cold-start relay+circuit has not been built yet. the top 270r 5w resistors emit around 240f/115c at idle and lower 100r 3w are comfortably less than that (didnt do a focused measurement on them).

front end triple zener is 160v x3 even though the sim has been update to a newer pending revision (comparing 500v vs 520v front end limiters). the 200-12k-200 for the back end limiter will be revised to 420v-440v range since this system will be dialed-in to handle 130v mains 24/7/365 or at least thats the design plan.

as you can see the inverted v-doubler has been replaced with series "film" cap feeding a bridge rectifier.

the 11+33 zener regulator will be replaced with either 47v or 51v 5w single. the zener's in the circuit have had their resistance specs matched to the data sheet. the sim normally uses a low level/blank template for diodes and various components. can change capacitor esr if needed.

the tube resistor equivalents have been adjusted so that the front-end / b+ / b1 sim numbers match bench tests (reverse engineering the sim for improved accuracy).

sim data file attached.

 

[sae-7125]

a few changes to the amp circuit have occurred.

the signal input side 4.7k has been bumped up to 10k. might overhaul the four components completely, havent decided yet.

the 47k new replacement for the 25k measured carbon comp posted recently might be a source of annoyance. low level speaker oscillation (1-2hz) is noticeable plus power off mild pop and woofer excursion (pulling at power off transition). exploring 33k 9w solution (100k 3w parallel x3). the new 47k 1w would hit around 220f/104c after ~10 minutes of idle. after installing the 33k mod the temp is down to 160f/71c and woofer pop/noise greatly reduced. this increases the current for v1 second triode to 5.1ma which is higher than the other three triodes in the front end amp circuit.

 

[sae-7125]

measuring across one of the output tubes (g1- to cathode+) showed very unstable bias voltage reaction i.e. a significant frequency amplitude was present. after adjusting to 33k the bias voltage is much more stable. tried swapping v1 and v2 with newer glass and same result prior to the 33k adjustment.

more focus needed to reduce the unwanted noise/oscillation to zero or close as possible to zero.

although the amp output might be limited to around 10w, it can still rock the house/bench speakers. healthy bass response, we are not talk about bose wave radio bass either i.e. you can sense the presence of low frequency information but nothing beyond that.

a fun test track to rattle the walls/windows/dishes/neighbors-windows is:

https://soundcloud.com/skeewiff%2Fskeewiff-space-ghetto-waltz

*for those with headphones that can handle real bass or computer connected to the home/big system.

its nice that the warehouse walls are block/concrete including floor and ceiling. and the neighbors are usual gone after 7pm.

 

[sae-7125]

the ladder bias core 510//510 is being adjusted lower. currently 255//1.5k = 218, might do 250r 25w//1k 5w = 200.

possibly playing with cathode cap values and sonic flavor reaction soon but its somewhat trivial until the next batch of permanent parts arrive. this system will be tuned electrically for 6L6 and 0.9a heater spec bottles. the 6.3v txf tap can handle 1.3a each with the kt66, so a lower heater consumption spec will improved system service life.

doing more with less can be enjoyable.

 

[sae-7125]

figuring out mounting locations for the caps and upcoming batch of thick film resistors (chassis mountable)

 

[gadget73]

excellent question and observation.

another experiment to evaluate circuit response. went with 1uf instead of 20 since its also easier to source 1uf film caps at the needed voltage, not quite dime-a-dozen but still cost effective. basic sonic testing so far shows no lack of bass with the 1+1 parallel bench speakers. even going from 100uf to 47uf for the cathode caps didnt impact bass response in a bad way.

trying to do a full-film renovation/redesign for improved longevity.

its not so much lack of bass as lack of stability. Pulse the input with about a 1v DC and see what it does. If it settles rapidly you're good. If it bounces like a car with bad shocks, its not happy. A multimeter set for diode check is a handy source for voltage in this range. If there isn't enough cap to decouple the stages fully they can interact in undesirable ways.

The lower value cathode cap will drop the gain at the output stage at low frequencies and the feedback loop is compensating, You won't hear it, but the distortion will tend to increase with a smaller cap because of this. Higher output impedance too, so the bass can get looser.

 

[sae-7125]

Pulse the input with about a 1v DC and see what it does. If it settles rapidly you're good. If it bounces like a car with bad shocks, its not happy.

hmm, further testing definitely needed. may end up adjusting that triodes (47k reduced to 33k) cap from 1uf to 4.7uf and reevaluate the results.

The lower value cathode cap will drop the gain at the output stage at low frequencies and the feedback loop is compensating,

it will be interesting to do a frequency sweep once the b+ board is finalized. possible fine tuning of the three resistors that feed v1 and v2 (next to 20uf original schematic caps).

 

[sae-7125]

top down view of bias ladder and the new g2 resistors waiting for a regulated power source

eventually the empty ground wire going between ladder and side of the chassis will be removed.

 

[sae-7125]

mounted the large filter caps. adjusting the pre-rectifier cap value.

started with 140+140v for the g2 regulator.

 

[sae-7125]

group photo with some r&d numbers

front end: 436v
b+: 383v
g2: 257v (under voltage vs zener regulator value, 62k too high and suffocating the sub-circuit)

reason for the parallel blob of caps in front of the front-end inverse v-doubler board... some of us werent paying attention to the board mounted parallel cap total i.e. 470nf//220nf//220nf = around 910nf +/- variance and not the 820nf target. mental fog can only excuse so much before self-inflicted blunder parameters are achieved +/- d'oh ^ cubed.

D'oh! - Wikipedia

en.wikipedia.org

testing at a lower voltage, high enough for close to full heater activity and yet lower b+ starting stress.

 

[sae-7125]

i suppose its time to spice things up with some r&d thermals


with the previous board design using the npn filter and inverted v-doubler

 

[sae-7125]

0.22r 5w is the hot spot in this view, helping to reduce 6.3v from over-voltage

 

[sae-7125]

revisiting the basics and possibly fine tuning v1/v2 circuits.

Coupling VS Decoupling VS Bypass Capacitor in Electronic Circuits

Capacitors are everywhere in electronics, but not all of them do the same job. Even though decoupling, bypass, and coupling capacitors often look identical

www.etechnophiles.com

the current circuit version being tested has received some +/- resistor adjustments to evaluate better or worse idle stability. going from 2.2k v2 cathode to 1.5k made the low frequency woofer motor-boating far worse so went in the other direction to 3k. changing the two 470k in that same area down to 120k also improved the twitchy idle bias measured across g1 and cathode (outputs).

bumping up 1uf at v1 second triode to 4.7uf (originally 20uf multi-cap) had little to no effect on bias stability even with rca input shorted.

stabilizing idle bias and reducing power off pop/woofer pull-in would be nice. more adjustment needed. the 47k feeding v1.1 anode and v1.2 grid is sitting at 33.33k which improved idle bias stability noticeably, might install 30k or low as 28k even though already calculating 5.1ma in that area.

 

[sae-7125]

bias ladder center resistor(s) hot spot. currently two 510r 7w in parallel.

 

[sae-7125]

still considering a homegrown smps with the parameters posted earlier in the thread.

SEPIC vs Cuk vs Zeta Converter

My understanding is that SEPICs, Cuks, and Zeta converters all provide a stable output voltage despite an input voltage that can vary either above or below what is desired. The SEPIC and Zeta maint...

electronics.stackexchange.com

 

[sae-7125]

hot spot is the original 47k feeding v1.1 anode and v1.2 grid. the resistor that measured 25k out of circuit shown earlier.

future thermals will be a little more up to date since there is a backlog of notes to sort through.

 

[sae-7125]

closer look at updates and cleaning up the v1/v2 front end. still lots of carbon comp to replace when this photo was taken. currently 99% carbon comp has been removed from the circuit.


one ulterior motive for making the w-5m truly 130v tolerant is that the warehouse mains numbers run on the high side of the voltage spectrum. the biz suite has been known to hit 132v and who knows what the full daily range might be, need a data-logger to watch the 30 day average.

 

[sae-7125]

watching the numbers, b+ 417v, g2 262v and cathode-g1 -18.6v bias


slowly waking up the amp design with higher idle numbers.

using an sae preamp to feed higher signal since a passive preamp can only go so far.

 

[sae-7125]

30uf caps mounted and the rectifier change over is being tested. still looks like a radio shack store exploded inside the chassis.

 

[sae-7125]

experimenting with changing the 47k to 33.33k 9w