Hello Bear,
I'm just a lowly builder, not an engineer except in the computer world. And even there I'm just MCSE. (Micro$oft Windoz) But my job is to build the company's RAID servers, clusters, firewalls, mail server, clients, Novell server, Linux CVS and Web server and maintain both the hardware and the software. Always too busy to bother with Cisco Certs. Got to do some serious testing before a new server goes up live. Real picky boss. So I believe in serious testing. But this is the world of 1's and 0's.
You make an interesting point about testing that has always been unclear to me.
"test audio output transformers with real world audio loads and power levels"
I can successfully test an amp with a dummy load, tone signals etc., etc., et cetera and come up with some pretty good answers. Soon's I hook up a speaker to it and play some real music, everything seems to fall apart. The music keeps changing everything, the speaker impedance jumps all over the place, the power on the rails keep shifting with the music, and I end up with not much information that makes any sense. And that's with just my shopdog speaker that I don't care if it blows up.
From your many queries to Susan it would seem that you know just about everything there is to know about the audio world.
So how is one to test in the "real world loads and power levels" when that ultimately means listening to music?
After all the careful testing, I always seem to end up using my ears. No 1's and 0's available here.
I remember Susan expressing that same sentiment albeit in a somewhat different phrasing.
As for "to make it practical at all there needs to be considerable interest", hardly any of my co-workers find my room full of gear practical. But they are just computer geeks who want it to sound "good" to them. However, I am very serious about knowing how to test in the real world.
Hoping you can enlighten us all.
Prosit
I'm just a lowly builder, not an engineer except in the computer world. And even there I'm just MCSE. (Micro$oft Windoz) But my job is to build the company's RAID servers, clusters, firewalls, mail server, clients, Novell server, Linux CVS and Web server and maintain both the hardware and the software. Always too busy to bother with Cisco Certs. Got to do some serious testing before a new server goes up live. Real picky boss. So I believe in serious testing. But this is the world of 1's and 0's.
You make an interesting point about testing that has always been unclear to me.
"test audio output transformers with real world audio loads and power levels"
I can successfully test an amp with a dummy load, tone signals etc., etc., et cetera and come up with some pretty good answers. Soon's I hook up a speaker to it and play some real music, everything seems to fall apart. The music keeps changing everything, the speaker impedance jumps all over the place, the power on the rails keep shifting with the music, and I end up with not much information that makes any sense. And that's with just my shopdog speaker that I don't care if it blows up.
From your many queries to Susan it would seem that you know just about everything there is to know about the audio world.
So how is one to test in the "real world loads and power levels" when that ultimately means listening to music?
After all the careful testing, I always seem to end up using my ears. No 1's and 0's available here.
I remember Susan expressing that same sentiment albeit in a somewhat different phrasing.
As for "to make it practical at all there needs to be considerable interest", hardly any of my co-workers find my room full of gear practical. But they are just computer geeks who want it to sound "good" to them. However, I am very serious about knowing how to test in the real world.
Hoping you can enlighten us all.
Prosit
Dear Bear,
Thank you for your further thoughts on this matter
I am sorry but I am somewhat confused here. My amplifier runs the mosfets as followers (with a gain of a smidgen under unity) and the transformer generates the negative power rail.
Both devices remain ON at all times in normal operation, and thus remain in Class A (i.e. not being switched off).
If I was using the mosfets as amplifiers then yes, each would turn off for part of the other's cycle, which would be dependent on the bias point.
Of course - I will immediately go to the attic and start reading my 1930's, 40's, 50 and 60's issues (yes, I do have them - including the issue for the outbreak of war and I have the slip to hand to the newsagent to reserve future copies of forthcoming issues).
If that fails I can always go over to the IEE library (I am a MIEE after all and what is institution membership for if one cannot avail oneself of some advantage). Luncheon for members is also available at a reasonable cost and the views out across the Thames can be very pleasant Only down side is that I will have to go during the week as The Institution is not normally open at weekends.
Please enlighten us to these references and perhaps a scan of the appropriate pages (which I am sure is within the "fair use" copyright clause).
Ah, you are showing my weak and lowly position as I have to admit to not having a variable low impedance non inductive load. I feel great shame and public humiliation.
Yes, I did say about running low impedance loads, by using different output winding arrangements - not on the straight 1+1:1 configuration.
I understand about op-amps - little beetle like things that scurry about on matrix boards or PCBs. I find the ceramic ones with gold legs particularly attractive, the black matte plastic ones are rather ho-hum.
Ah, perhaps all is not lost after all
... pause ...
okay, with the MRF148 setup:
At 1.005 kHz, setting for 2 volts baseline:
2.004 volts ac-rms into 8.016 ohms (3.6 uH) load
1.573 volts ac-rms into 4.028 ohms (2.9 uH) load
2.855 volts ac-rms into open circuit (100 k ohms).
I am sorry, but this I also do not understand
I look at the impedance curves for the wide-range drivers that I use, or am interested in, and even at a lowly 20 kHz they have impedances of around 25 ohms, and rising. I would estimate from the curves that they will hit 50 ohm impedance range at or soon after 100 kHz, which is audio these days.
Or what are we doing with 24 bit, 192 kHz audio ADC/DACs?
The output stage of a power amp can potentially operate into the tens or even hundreds of megahertz range. You may not hear it from the speakers, but that is not the point.
To maintain fidelity for a 100 kHz audio bandwidth signal the audio chain itself should have a (minimum) bandwidth of 1 MHz. This is audio at RF frequencies and it behooves a person to know what her amplifier is going to do here just as much as between 10 Hz and 20 kHz.
And if RF is not important then why was there another DIY audio thread about GSM interference and how to solve it - and this is at 800/900 or 1800 MHz?
Unfortunately it is incumbent upon anyone who is building any form of electronics to take reasonable steps to ensure that they do not cause EMC/RFI interference, DIYer or professional.
I am sure the FCC would say the same thing, if not more so as I believe that the FCC are even more concerned with emissions than here in Europe.
Most hobby electronics is below the radar of the powers that be, but this does not mean open house.
I already have a pair of classic QUAD electrostatic loudspeakers and they are HUGE compared to the overall room size (which we have to live in - it is a living room after all). As I have previously mentioned, I have lent them as I don't have the space here. (We live in a one bed apartment - the kitchen is 8 by 8 feet.) Electrostatics simply don't work well in a room of this size, they need to be at least six feet from the rear wall which would put them in the middle of the room!
The quad filar windings are easy. The 0.8 mm wire is thick enough to be mostly self supporting and I just hold the four wires in one hand and turn the bobbin in the other. Oops, sorry that would be 20 AWG for you.
The difficult one is the input transformer, and that has lots of turns of fine wire and really needs a proper coil winder to get even layers and matched turns. And for two off it is easier to buy a transformer from Sowter - it's already designed, will be built to a certifiable standard, and has a part number.
I thank you for your post and I hope my answers are not inappropriate.
Best wishes,
Susan.
Thank you for your further thoughts on this matter
bear said:
I am sorry but I am somewhat confused here. My amplifier runs the mosfets as followers (with a gain of a smidgen under unity) and the transformer generates the negative power rail.
Both devices remain ON at all times in normal operation, and thus remain in Class A (i.e. not being switched off).
If I was using the mosfets as amplifiers then yes, each would turn off for part of the other's cycle, which would be dependent on the bias point.
Of course - I will immediately go to the attic and start reading my 1930's, 40's, 50 and 60's issues (yes, I do have them - including the issue for the outbreak of war and I have the slip to hand to the newsagent to reserve future copies of forthcoming issues).
If that fails I can always go over to the IEE library (I am a MIEE after all and what is institution membership for if one cannot avail oneself of some advantage). Luncheon for members is also available at a reasonable cost and the views out across the Thames can be very pleasant
Only down side is that I will have to go during the week as The Institution is not normally open at weekends. Please enlighten us to these references and perhaps a scan of the appropriate pages (which I am sure is within the "fair use" copyright clause).
Ah, you are showing my weak and lowly position as I have to admit to not having a variable low impedance non inductive load. I feel great shame and public humiliation.
Yes, I did say about running low impedance loads, by using different output winding arrangements - not on the straight 1+1:1 configuration.
I understand about op-amps - little beetle like things that scurry about on matrix boards or PCBs. I find the ceramic ones with gold legs particularly attractive, the black matte plastic ones are rather ho-hum.
Ah, perhaps all is not lost after all
... pause ...
okay, with the MRF148 setup:
At 1.005 kHz, setting for 2 volts baseline:
2.004 volts ac-rms into 8.016 ohms (3.6 uH) load
1.573 volts ac-rms into 4.028 ohms (2.9 uH) load
2.855 volts ac-rms into open circuit (100 k ohms).
I am sorry, but this I also do not understand
I look at the impedance curves for the wide-range drivers that I use, or am interested in, and even at a lowly 20 kHz they have impedances of around 25 ohms, and rising. I would estimate from the curves that they will hit 50 ohm impedance range at or soon after 100 kHz, which is audio these days.
Or what are we doing with 24 bit, 192 kHz audio ADC/DACs?
The output stage of a power amp can potentially operate into the tens or even hundreds of megahertz range. You may not hear it from the speakers, but that is not the point.
To maintain fidelity for a 100 kHz audio bandwidth signal the audio chain itself should have a (minimum) bandwidth of 1 MHz. This is audio at RF frequencies and it behooves a person to know what her amplifier is going to do here just as much as between 10 Hz and 20 kHz.
And if RF is not important then why was there another DIY audio thread about GSM interference and how to solve it - and this is at 800/900 or 1800 MHz?
Unfortunately it is incumbent upon anyone who is building any form of electronics to take reasonable steps to ensure that they do not cause EMC/RFI interference, DIYer or professional.
I am sure the FCC would say the same thing, if not more so as I believe that the FCC are even more concerned with emissions than here in Europe.
Most hobby electronics is below the radar of the powers that be, but this does not mean open house.
I already have a pair of classic QUAD electrostatic loudspeakers and they are HUGE compared to the overall room size (which we have to live in - it is a living room after all). As I have previously mentioned, I have lent them as I don't have the space here. (We live in a one bed apartment - the kitchen is 8 by 8 feet.) Electrostatics simply don't work well in a room of this size, they need to be at least six feet from the rear wall which would put them in the middle of the room!
The quad filar windings are easy. The 0.8 mm wire is thick enough to be mostly self supporting and I just hold the four wires in one hand and turn the bobbin in the other. Oops, sorry that would be 20 AWG for you.
The difficult one is the input transformer, and that has lots of turns of fine wire and really needs a proper coil winder to get even layers and matched turns. And for two off it is easier to buy a transformer from Sowter - it's already designed, will be built to a certifiable standard, and has a part number.
I thank you for your post and I hope my answers are not inappropriate.
Best wishes,
Susan.
Hi,
Well, I can have a quality three course meal and coffee for not much more than the price of a coffee and a piece of cake at starbucks.
The food has improved a lot in the last four years.
Plus I can fit in a visit the Library - which it would appear is something I need to do immediately.
As to bandwidth, it's to preserve phase (shift starts a decade below the top end response) and to ensure (in a feedback amplifier) that one doesn't run out of gain - else the feedback bit doesn't work).
My thoughts on the matter anyway.
Best wishes,
Susan.
traderbam said:
Well, I can have a quality three course meal and coffee for not much more than the price of a coffee and a piece of cake at starbucks.
The food has improved a lot in the last four years.
Plus I can fit in a visit the Library - which it would appear is something I need to do immediately.
As to bandwidth, it's to preserve phase (shift starts a decade below the top end response) and to ensure (in a feedback amplifier) that one doesn't run out of gain - else the feedback bit doesn't work).
My thoughts on the matter anyway.
Best wishes,
Susan.
acenovelty said:
The idea that I bring up is to simply have reasonable levels of core saturation/use when you test. Transformers will have apparently better response specs generally speaking when there is less signal.
The issue of an amp "coming apart" with a real speaker load is another matter. With a *feedback* type amp, that is fairly predictable, as the *reactive load* presents a changing phase angle with frequency to the amp, and the feedback tries to correct that with an amplitude adjustment, which never actually works exactly right. The *better* the inherent amp is, the more effective the feedback is. Tube amps generate all sorts of nasty problems with feedback due to the limitations of the xfmr which then are trying to be corrected by the feedback...
Anyhow, with a non-feedback amp, the changing impedance load and the reactive elements in the xover & the speaker itself are not simply resistive, and present the same problems as were mentioned before. The resulting issues are much the same, but the results as far as the amplifiers performance are somewhat different.
For testing, you can always build a simple passive circuit that "emulates" a speaker load... but for most bench tests the use of regular resistors will do fine.
And no, I don't know "everything about audio" by any stretch of the imagination... there are quite a number of posters on this forum that trump my knowledge by a wide margin on certain topics...
Susan-Parker said:Dear Bear,
Thank you for your further thoughts on this matter
I am sorry but I am somewhat confused here. My amplifier runs the mosfets as followers (with a gain of a smidgen under unity) and the transformer generates the negative power rail.
Both devices remain ON at all times in normal operation, and thus remain in Class A (i.e. not being switched off).
If I was using the mosfets as amplifiers then yes, each would turn off for part of the other's cycle, which would be dependent on the bias point.
Measure the current through ur mosfets as they sit there.
Put a small resistor or ammeter in series and see where they are.
You don't need non-inductive resistors for this measurement. You're not going to be testing up to a high enough frequency where that makes any difference at all. In fact ur not testing wrt frequency.
Regular old paralleled 10 watt resistors will work fine.
I understand about op-amps - little beetle like things that scurry about on matrix boards or PCBs. I find the ceramic ones with gold legs particularly attractive, the black matte plastic ones are rather ho-hum.
Ah, perhaps all is not lost after all
Yeah I like the gold and ceramic ones...
A few more data points would be useful...
looks like a minor mismatch at 8 ohms... since open circuit shows
2.855 volts... less at 4 ohms... which equals:
0.614w @ 4ohms
0.502w @ 8ohms
What's it doing at 6, 2 and one ohm? 16 ohms?
Yeah, well 25 ohm drivers are kind of unusual these days... not
unheard of though. Not what most folks will be using.
As far as the hi rate dacs? We're waiting for decent source material?
No it won't. Nor should it. Nor will the layout permit such operation. Most, not all solid state direct coupled amps have a small inductor at the ouput to prevent the amp seeing a super low Z reactance load at HF, causing the amp to see a short. All input circuits and feedback circuits have components to limit the bandwidth of the amp. The sole exception to this in commercial amplifiers is the Spectral line which due to intentional design, careful circuit layout is capable of bandwidths to 1 Mhz for the power amps and 2 Mhz for the preamps.
Regardless, this is a *spectacularly* bad design idea - making the amp that wide of a bandwidth. Your AM broadcast transmitters in the 50kw range operate starting at ~500kHz. The last thing any manufacturer wants is for someone living next door to a broadcaster to be amplifying that!
Your transformer coupled amp will pass NOTHING substantive at ~500kHz due to the self inductance of the xfmr.
If you used amorphous cores and extreme winding techniques you might find that you can push the ~250kHz limit slightly.
In theory you are correct. In practice there are as of yet no 100kHz. audio signals. No audio chain has this bandwidth at this time - the Spectral components do as far as a preamp & amp - there are no speakers that will do this. The best you can do there is ~>10hz. - ~100kHz., and that is a full out price no object system! And there is no source that comes close.
No one really is going to get or need that sort of bandwidth today, if ever. Sure, look at it, but that's not the main issue for practical designs. Take more time to look at the spectra of distortion.
dunno - didn't read it.
You are worrying unecessarily...
Not quite so easy on an automatic machine...
Straight from the horses, er......... Bears mouth.
"BEAR Labs designs and builds equipment that is technically superior and unique.
In most audio gear, especially speakers, there are no definitive and unequivocally “perfect” designs. So there are always sets of choices and compromises that engineers and designers must make.
Those who rely solely upon “engineering and science” often make a simple mistake of believing that measurements and/or equations can truly account for all factors in design.
Those who rely solely upon “engineering and science” often make a simple mistake of believing that measurements and/or equations can truly account for all factors in design.
What is audible? What makes a piece of equipment good? How good does a given piece of equipment have to be? Can things like parts and wires actually make an audible difference?
The difference between good, very good, and nearly perfect gear is the central issue! After all, if this audible difference did not exist and diminish listening pleasure then there would be no incentive to apply the technology, materials and expense involved in building high quality equipment!
Describing these crucial differences is difficult and almost impossible to achieve in words, but easy to identify once you compare BEAR Labs and competing equipment - by simply listening!
But, let's not avoid the question. Simply stated, the difference is clarity. Clarity in its simplest terms means that the superior equipment reproduces all of the harmonics (the subtle differences in timbre) so correctly and precisely that your brain recognizes exactly what it is hearing more quickly and accurately. In simple terms that is the difference.
In essence, there is no "standard" by which one can assemble a system and know, a priori, how it will perform and how it will sound.
Keep in mind when we are thinking about audiophile systems that the ultimate comparison has to made to real live sound. Audiophile systems today can at best recreate a crude approximation of live sound. Even so, the less “electronic” or “mechanical” the sound of said system, the better.
So, when you acquire BEAR Labs equipment you will have purchased equipment that will not only last almost forever, but is designed to perform so well as to meet your needs for years and years to come."
All others need not apply.
Prosit
"BEAR Labs designs and builds equipment that is technically superior and unique.
In most audio gear, especially speakers, there are no definitive and unequivocally “perfect” designs. So there are always sets of choices and compromises that engineers and designers must make.
Those who rely solely upon “engineering and science” often make a simple mistake of believing that measurements and/or equations can truly account for all factors in design.
Those who rely solely upon “engineering and science” often make a simple mistake of believing that measurements and/or equations can truly account for all factors in design.
What is audible? What makes a piece of equipment good? How good does a given piece of equipment have to be? Can things like parts and wires actually make an audible difference?
The difference between good, very good, and nearly perfect gear is the central issue! After all, if this audible difference did not exist and diminish listening pleasure then there would be no incentive to apply the technology, materials and expense involved in building high quality equipment!
Describing these crucial differences is difficult and almost impossible to achieve in words, but easy to identify once you compare BEAR Labs and competing equipment - by simply listening!
But, let's not avoid the question. Simply stated, the difference is clarity. Clarity in its simplest terms means that the superior equipment reproduces all of the harmonics (the subtle differences in timbre) so correctly and precisely that your brain recognizes exactly what it is hearing more quickly and accurately. In simple terms that is the difference.
In essence, there is no "standard" by which one can assemble a system and know, a priori, how it will perform and how it will sound.
Keep in mind when we are thinking about audiophile systems that the ultimate comparison has to made to real live sound. Audiophile systems today can at best recreate a crude approximation of live sound. Even so, the less “electronic” or “mechanical” the sound of said system, the better.
So, when you acquire BEAR Labs equipment you will have purchased equipment that will not only last almost forever, but is designed to perform so well as to meet your needs for years and years to come."
All others need not apply.
Prosit
Ok, here's a suggestion from a newbie:
Since we have plenty of people with extensive experience and expertise in the field of audio design and manufacturing.....
Instead of peppering Susan with a scattergraph of quasi-random request, why don't we come up with a coherent 1 thru 10 list of tests to perform on her amp along with directions for the easiest way to perform them.
And for the sake of us newbies who follow these threads closely, could those "in the know" please include explanations why these tests are pertinent and how they relate to other matters.
We've already had some great posts from bear, John Curl, AKSA, Circlotron, Graham Maynard, Mikeks, etc.
So what do you guys think?
Would it be possible to compile a list of tests in a logical order that would allow others to understand how(and how well) this amp works?
And also teach the newbies a few things about testing and design.
Since we have plenty of people with extensive experience and expertise in the field of audio design and manufacturing.....
Instead of peppering Susan with a scattergraph of quasi-random request, why don't we come up with a coherent 1 thru 10 list of tests to perform on her amp along with directions for the easiest way to perform them.
And for the sake of us newbies who follow these threads closely, could those "in the know" please include explanations why these tests are pertinent and how they relate to other matters.
We've already had some great posts from bear, John Curl, AKSA, Circlotron, Graham Maynard, Mikeks, etc.
So what do you guys think?
Would it be possible to compile a list of tests in a logical order that would allow others to understand how(and how well) this amp works?
And also teach the newbies a few things about testing and design.
Dear All,
Test with IRFP150s at 1.005 kHz - normalized at 2.006 volts ac-RMS
Load - volts
4.028 - 1.660
8.024 - 2.006
16.45 - 2.262
33.23 - 2.357
65.79 - 2.419
100.1 - 2.442
100K - 2.488
I am not testing below 4 ohms as the transformer secondaries would be configured differently.
Best wishes,
Susan.
P.S.
The speaker impedances I was talking about being in the 50 ohms range above 100 kHz are for nominal 8 ohm drivers - not something special or exotic.
The quiescent current for the amplifier with 225 to 240 mV source to ground via the transformer is around 330 mA.
I already have a 0.1 ohm resistor in circuit, plus a 6 Amp fuse (after the smoothing cap).
I don't consider 35 plus watts into 8 ohms as an unreasonable level of core saturation/use, I get 8 watts at 10 Hz.
Susan.
Test with IRFP150s at 1.005 kHz - normalized at 2.006 volts ac-RMS
Load - volts
4.028 - 1.660
8.024 - 2.006
16.45 - 2.262
33.23 - 2.357
65.79 - 2.419
100.1 - 2.442
100K - 2.488
I am not testing below 4 ohms as the transformer secondaries would be configured differently.
Best wishes,
Susan.
P.S.
The speaker impedances I was talking about being in the 50 ohms range above 100 kHz are for nominal 8 ohm drivers - not something special or exotic.
The quiescent current for the amplifier with 225 to 240 mV source to ground via the transformer is around 330 mA.
I already have a 0.1 ohm resistor in circuit, plus a 6 Amp fuse (after the smoothing cap).
I don't consider 35 plus watts into 8 ohms as an unreasonable level of core saturation/use, I get 8 watts at 10 Hz.
Susan.
Susan's amplifier, effectively a transformer driven tube design using mosfets, will give marvellous sound. I know; I've heard amps just like them, made with mosfets, and designed one myself using a tube driver. Right now I'm working on an up-market, Class A differential amp using similar topology and a tube front end. I fully expect it to sound sensational.
I believe Susan has given all the necessary data any reasonable person would want in close technical description of her design.
She has offered to have transformers made.
She has given lucid description of the sonics, which precisely corresponds with my own experience.
Now, we have a spray of complex, IUC-related questions, demonstrating profound knowledge of 'the problem'.
Hey guys, hello, why not build the !@#$ thing? Have a listen, perhaps? Most of life's seminal experiences are nothing but words until we actually try them
Thanks Susan!
Cheers,
Hugh
I believe Susan has given all the necessary data any reasonable person would want in close technical description of her design.
She has offered to have transformers made.
She has given lucid description of the sonics, which precisely corresponds with my own experience.
Now, we have a spray of complex, IUC-related questions, demonstrating profound knowledge of 'the problem'.
Hey guys, hello, why not build the !@#$ thing? Have a listen, perhaps? Most of life's seminal experiences are nothing but words until we actually try them
Thanks Susan!
Cheers,
Hugh
I would still prefer a simple chip amp with a RLC series network in series with its output [C=680uF cheap non-linear electrolytic, R=2 ohm low power so heat could change its value, L=30uH w/cheap non-linear iron powder core]. It should produce the same 'tubish' colouration due to loudspeaker impedance fluctuations [including signal-dependent impedance changes] and non-linear output impedance but at a fraction of the cost/space requirements
Susan, and all,
I would be curious how would a cascode transistor on top on the FET's work in your amplifier.
This way the non-linear input capacitance wont be so source impedance critical.
Just a curious idea in such a design and how it could work and sound, what do you folks think?
Regards
Michael
I would be curious how would a cascode transistor on top on the FET's work in your amplifier.
This way the non-linear input capacitance wont be so source impedance critical.
Just a curious idea in such a design and how it could work and sound, what do you folks think?
Regards
Michael
I've a much better idea; build one and listen to it.
It's the ONLY thing that matters in my book and the only thing that can tell you how it sounds.
I quite fancy having a go myself, it's just so different from everything I've seen before, which makes me very curious.
Andy.
Edit: Just read Hugh's reply, I agree 100%
Eva said:I would still prefer a simple chip amp with a RLC series network in series with its output [C=680uF cheap non-linear electrolytic, R=2 ohm low power so heat could change its value, L=30uH w/cheap non-linear iron powder core]. It should produce the same 'tubish' colouration due to loudspeaker impedance fluctuations [including signal-dependent impedance changes] and non-linear output impedance but at a fraction of the cost/space requirements
Ah, but one then misses out on all the custom hot rod extras like shiny chrome transformer lids and big fat smoothing electrolytics (valve subsitutes). And the latter could even be fitted inside Perspex tubes with blue LEDs to make the whole thing glow and match those swanky see thru computer case with neon effect fans and cables - or the "crusing lights" on the underside of one's truck.
Perceived value is all, if it ain't fitted with a Chevvy big block n' chromed head to pipe then it ain't worth a dime.
BW,
Susan.
P.S.
Don't forget the nitrous conversion and the chute...